diff --git a/config/config.yaml b/config/config.yaml
deleted file mode 100644
index 1e524ed5f..000000000
--- a/config/config.yaml
+++ /dev/null
@@ -1,4 +0,0 @@
-defaults:
-  - datasets: gcp
-  - step/session: dataproc
-  - step: ???
diff --git a/config/datasets/gcp.yaml b/config/datasets/ot_gcp.yaml
similarity index 100%
rename from config/datasets/gcp.yaml
rename to config/datasets/ot_gcp.yaml
diff --git a/config/ot_config.yaml b/config/ot_config.yaml
new file mode 100644
index 000000000..7f28a58d6
--- /dev/null
+++ b/config/ot_config.yaml
@@ -0,0 +1,5 @@
+defaults:
+  - config
+  - datasets: ot_gcp
+  - _self_
+  - override step/session: dataproc
diff --git a/config/step/finngen_sumstat_preprocess.yaml b/config/step/finngen_sumstat_preprocess.yaml
deleted file mode 100644
index 319e7af63..000000000
--- a/config/step/finngen_sumstat_preprocess.yaml
+++ /dev/null
@@ -1,3 +0,0 @@
-_target_: otg.finngen_sumstat_preprocess.FinnGenSumstatPreprocessStep
-raw_sumstats_path: ???
-out_sumstats_path: ???
diff --git a/config/step/gwas_catalog_sumstat_preprocess.yaml b/config/step/gwas_catalog_sumstat_preprocess.yaml
deleted file mode 100644
index 29d3486e8..000000000
--- a/config/step/gwas_catalog_sumstat_preprocess.yaml
+++ /dev/null
@@ -1,3 +0,0 @@
-_target_: otg.gwas_catalog_sumstat_preprocess.GWASCatalogSumstatsPreprocessStep
-raw_sumstats_path: ???
-out_sumstats_path: ???
diff --git a/config/step/ld_index.yaml b/config/step/ld_index.yaml
deleted file mode 100644
index 86f595076..000000000
--- a/config/step/ld_index.yaml
+++ /dev/null
@@ -1,6 +0,0 @@
-_target_: otg.ld_index.LDIndexStep
-
-session:
-  start_hail: true
-
-ld_index_out: ${datasets.ld_index}
diff --git a/config/step/colocalisation.yaml b/config/step/ot_colocalisation.yaml
similarity index 74%
rename from config/step/colocalisation.yaml
rename to config/step/ot_colocalisation.yaml
index 2a4062a56..97cfa1bea 100644
--- a/config/step/colocalisation.yaml
+++ b/config/step/ot_colocalisation.yaml
@@ -1,4 +1,4 @@
 _target_: otg.colocalisation.ColocalisationStep
-credible_set_path: ${datasets.credible_set}
+credible_set_path: ${datasets.study_locus}
 study_index_path: ${datasets.study_index}
 coloc_path: ${datasets.colocalisation}
diff --git a/config/step/eqtl_catalogue.yaml b/config/step/ot_eqtl_catalogue.yaml
similarity index 82%
rename from config/step/eqtl_catalogue.yaml
rename to config/step/ot_eqtl_catalogue.yaml
index 04a958993..141c94ee3 100644
--- a/config/step/eqtl_catalogue.yaml
+++ b/config/step/ot_eqtl_catalogue.yaml
@@ -1,4 +1,6 @@
-_target_: otg.eqtl_catalogue.EqtlCatalogueStep
+defaults:
+  - eqtl_catalogue
+
 eqtl_catalogue_paths_imported: ${datasets.eqtl_catalogue_paths_imported}
 eqtl_catalogue_study_index_out: ${datasets.eqtl_catalogue_study_index_out}
 eqtl_catalogue_summary_stats_out: ${datasets.eqtl_catalogue_summary_stats_out}
diff --git a/config/step/finngen_studies.yaml b/config/step/ot_finngen_studies.yaml
similarity index 53%
rename from config/step/finngen_studies.yaml
rename to config/step/ot_finngen_studies.yaml
index 23b58c443..673e0f269 100644
--- a/config/step/finngen_studies.yaml
+++ b/config/step/ot_finngen_studies.yaml
@@ -1,2 +1,4 @@
-_target_: otg.finngen_studies.FinnGenStudiesStep
+defaults:
+  - finngen_studies
+
 finngen_study_index_out: ${datasets.finngen_study_index}
diff --git a/config/step/ot_finngen_sumstat_preprocess.yaml b/config/step/ot_finngen_sumstat_preprocess.yaml
new file mode 100644
index 000000000..ad0e93a09
--- /dev/null
+++ b/config/step/ot_finngen_sumstat_preprocess.yaml
@@ -0,0 +1,5 @@
+defaults:
+  - finngen_sumstat_preprocess
+
+raw_sumstats_path: ???
+out_sumstats_path: ???
diff --git a/config/step/gene_index.yaml b/config/step/ot_gene_index.yaml
similarity index 66%
rename from config/step/gene_index.yaml
rename to config/step/ot_gene_index.yaml
index 20de8dfbe..ce5971bf9 100644
--- a/config/step/gene_index.yaml
+++ b/config/step/ot_gene_index.yaml
@@ -1,3 +1,5 @@
-_target_: otg.gene_index.GeneIndexStep
+defaults:
+  - gene_index
+
 target_path: ${datasets.target_index}
 gene_index_path: ${datasets.gene_index}
diff --git a/config/step/gwas_catalog_ingestion.yaml b/config/step/ot_gwas_catalog_ingestion.yaml
similarity index 88%
rename from config/step/gwas_catalog_ingestion.yaml
rename to config/step/ot_gwas_catalog_ingestion.yaml
index 66fe37ce9..65606b7e4 100644
--- a/config/step/gwas_catalog_ingestion.yaml
+++ b/config/step/ot_gwas_catalog_ingestion.yaml
@@ -1,4 +1,6 @@
-_target_: otg.gwas_catalog_ingestion.GWASCatalogIngestionStep
+defaults:
+  - gwas_catalog_ingestion
+
 catalog_study_files: ${datasets.catalog_studies}
 catalog_ancestry_files: ${datasets.catalog_ancestries}
 catalog_associations_file: ${datasets.catalog_associations}
diff --git a/config/step/gwas_catalog_curation_update.yaml b/config/step/ot_gwas_catalog_study_curation.yaml
similarity index 79%
rename from config/step/gwas_catalog_curation_update.yaml
rename to config/step/ot_gwas_catalog_study_curation.yaml
index 979439153..eb6c0ec78 100644
--- a/config/step/gwas_catalog_curation_update.yaml
+++ b/config/step/ot_gwas_catalog_study_curation.yaml
@@ -1,4 +1,6 @@
-_target_: otg.gwas_catalog_study_curation.GWASCatalogStudyCurationStep
+defaults:
+  - gwas_catalog_study_curation
+
 catalog_study_files: ${datasets.catalog_studies}
 catalog_ancestry_files: ${datasets.catalog_ancestries}
 catalog_sumstats_lut: ${datasets.catalog_sumstats_lut}
diff --git a/config/step/gwas_study_inclusion.yaml b/config/step/ot_gwas_catalog_study_inclusion.yaml
similarity index 84%
rename from config/step/gwas_study_inclusion.yaml
rename to config/step/ot_gwas_catalog_study_inclusion.yaml
index 60916381b..8a560127e 100644
--- a/config/step/gwas_study_inclusion.yaml
+++ b/config/step/ot_gwas_catalog_study_inclusion.yaml
@@ -1,4 +1,6 @@
-_target_: otg.gwas_catalog_study_inclusion.GWASCatalogInclusionGenerator
+defaults:
+  - gwas_catalog_study_inclusion
+
 catalog_study_files: ${datasets.catalog_studies}
 catalog_ancestry_files: ${datasets.catalog_ancestries}
 catalog_associations_file: ${datasets.catalog_associations}
diff --git a/config/step/ot_gwas_catalog_sumstat_preprocess.yaml b/config/step/ot_gwas_catalog_sumstat_preprocess.yaml
new file mode 100644
index 000000000..d0c936807
--- /dev/null
+++ b/config/step/ot_gwas_catalog_sumstat_preprocess.yaml
@@ -0,0 +1,5 @@
+defaults:
+  - gwas_catalog_sumstat_preprocess
+
+raw_sumstats_path: ???
+out_sumstats_path: ???
diff --git a/config/step/ld_based_clumping.yaml b/config/step/ot_ld_based_clumping.yaml
similarity index 67%
rename from config/step/ld_based_clumping.yaml
rename to config/step/ot_ld_based_clumping.yaml
index eac354e4f..d25ca84b7 100644
--- a/config/step/ld_based_clumping.yaml
+++ b/config/step/ot_ld_based_clumping.yaml
@@ -1,4 +1,6 @@
-_target_: otg.ld_based_clumping.LdBasedClumpingStep
+defaults:
+  - ld_based_clumping
+
 study_locus_input_path: ???
 ld_index_path: ???
 study_index_path: ???
diff --git a/config/step/ot_ld_index.yaml b/config/step/ot_ld_index.yaml
new file mode 100644
index 000000000..70dc6b5ee
--- /dev/null
+++ b/config/step/ot_ld_index.yaml
@@ -0,0 +1,4 @@
+defaults:
+  - ld_index
+
+ld_index_out: ${datasets.ld_index}
diff --git a/config/step/locus_to_gene.yaml b/config/step/ot_locus_to_gene.yaml
similarity index 82%
rename from config/step/locus_to_gene.yaml
rename to config/step/ot_locus_to_gene.yaml
index 74c23b3d2..d055621ca 100644
--- a/config/step/locus_to_gene.yaml
+++ b/config/step/ot_locus_to_gene.yaml
@@ -1,8 +1,6 @@
-_target_: otg.l2g.LocusToGeneStep
+defaults:
+  - locus_to_gene
 
-session:
-  extended_spark_conf:
-    spark.dynamicAllocation.enabled: false
 run_mode: train
 wandb_run_name: null
 perform_cross_validation: false
diff --git a/config/step/pics.yaml b/config/step/ot_pics.yaml
similarity index 68%
rename from config/step/pics.yaml
rename to config/step/ot_pics.yaml
index 4a656a876..851c4ca06 100644
--- a/config/step/pics.yaml
+++ b/config/step/ot_pics.yaml
@@ -1,3 +1,5 @@
-_target_: otg.pics.PICSStep
+defaults:
+  - pics
+
 study_locus_ld_annotated_in: ???
 picsed_study_locus_out: ???
diff --git a/config/step/study_locus_overlap.yaml b/config/step/ot_study_locus_overlap.yaml
similarity index 79%
rename from config/step/study_locus_overlap.yaml
rename to config/step/ot_study_locus_overlap.yaml
index 02f9cd83d..acfcbdc94 100644
--- a/config/step/study_locus_overlap.yaml
+++ b/config/step/ot_study_locus_overlap.yaml
@@ -1,4 +1,6 @@
-_target_: otg.overlaps.OverlapsIndexStep
+defaults:
+  - overlaps
+
 study_locus_path: ${datasets.outputs}/credible_set
 study_index_path: ${datasets.outputs}/study_index
 overlaps_index_out: ${datasets.outputs}/study_locus_overlap
diff --git a/config/step/ukbiobank.yaml b/config/step/ot_ukbiobank.yaml
similarity index 74%
rename from config/step/ukbiobank.yaml
rename to config/step/ot_ukbiobank.yaml
index fc7d7fb19..a1dcb6d9f 100644
--- a/config/step/ukbiobank.yaml
+++ b/config/step/ot_ukbiobank.yaml
@@ -1,3 +1,4 @@
-_target_: otg.ukbiobank.UKBiobankStep
+defaults:
+  - ukbiobank
 ukbiobank_manifest: ${datasets.ukbiobank_manifest}
 ukbiobank_study_index_out: ${datasets.ukbiobank_study_index}
diff --git a/config/step/v2g.yaml b/config/step/ot_v2g.yaml
similarity index 88%
rename from config/step/v2g.yaml
rename to config/step/ot_v2g.yaml
index bdd003861..1ac6d2fbe 100644
--- a/config/step/v2g.yaml
+++ b/config/step/ot_v2g.yaml
@@ -1,10 +1,12 @@
-_target_: otg.v2g.V2GStep
+defaults:
+  - variant_to_gene
+
 variant_index_path: ${datasets.variant_index}
 variant_annotation_path: ${datasets.variant_annotation}
 gene_index_path: ${datasets.gene_index}
 vep_consequences_path: ${datasets.vep_consequences}
 liftover_chain_file_path: ${datasets.chain_37_38}
-intervals:
+interval_sources:
   andersson: ${datasets.anderson}
   javierre: ${datasets.javierre}
   jung: ${datasets.jung}
diff --git a/config/step/ot_variant_annotation.yaml b/config/step/ot_variant_annotation.yaml
new file mode 100644
index 000000000..5da76b398
--- /dev/null
+++ b/config/step/ot_variant_annotation.yaml
@@ -0,0 +1,4 @@
+defaults:
+  - variant_annotation
+
+variant_annotation_path: ${datasets.variant_annotation}
diff --git a/config/step/variant_index.yaml b/config/step/ot_variant_index.yaml
similarity index 76%
rename from config/step/variant_index.yaml
rename to config/step/ot_variant_index.yaml
index 88f84690b..1625c7126 100644
--- a/config/step/variant_index.yaml
+++ b/config/step/ot_variant_index.yaml
@@ -1,4 +1,6 @@
-_target_: otg.variant_index.VariantIndexStep
+defaults:
+  - variant_index
+
 variant_annotation_path: ${datasets.variant_annotation}
 credible_set_path: ${datasets.study_locus}
 variant_index_path: ${datasets.variant_index}
diff --git a/config/step/window_based_clumping.yaml b/config/step/ot_window_based_clumping.yaml
similarity index 50%
rename from config/step/window_based_clumping.yaml
rename to config/step/ot_window_based_clumping.yaml
index 55cb952c8..0e7e0e39b 100644
--- a/config/step/window_based_clumping.yaml
+++ b/config/step/ot_window_based_clumping.yaml
@@ -1,5 +1,6 @@
-_target_: otg.window_based_clumping.WindowBasedClumpingStep
+defaults:
+  - window_based_clumping
+
 summary_statistics_input_path: ???
 study_locus_output_path: ???
 inclusion_list_path: ???
-locus_collect_distance: null
diff --git a/config/step/session/dataproc.yaml b/config/step/session/dataproc.yaml
index 0c8b1208a..6ac641718 100644
--- a/config/step/session/dataproc.yaml
+++ b/config/step/session/dataproc.yaml
@@ -1,4 +1,5 @@
-_target_: otg.common.session.Session
+defaults:
+  - base_session
+
 spark_uri: yarn
-hail_home: /opt/conda/miniconda3/lib/python3.10/site-packages/hail
 write_mode: errorifexists
diff --git a/config/step/session/local.yaml b/config/step/session/local.yaml
deleted file mode 100644
index ac7f42591..000000000
--- a/config/step/session/local.yaml
+++ /dev/null
@@ -1,4 +0,0 @@
-_target_: otg.common.session.Session
-spark_uri: local[*]
-hail_home: ???
-write_mode: errorifexists
diff --git a/config/step/variant_annotation.yaml b/config/step/variant_annotation.yaml
deleted file mode 100644
index 1a8042ddb..000000000
--- a/config/step/variant_annotation.yaml
+++ /dev/null
@@ -1,6 +0,0 @@
-_target_: otg.variant_annotation.VariantAnnotationStep
-
-session:
-  start_hail: true
-
-variant_annotation_path: ${datasets.variant_annotation}
diff --git a/docs/development/airflow.md b/docs/development/airflow.md
index 9768bf961..b73ad614e 100644
--- a/docs/development/airflow.md
+++ b/docs/development/airflow.md
@@ -7,12 +7,12 @@ This section describes how to set up a local Airflow server which will orchestra
 - [Docker](https://docs.docker.com/get-docker/)
 - [Google Cloud SDK](https://cloud.google.com/sdk/docs/install)
 
-!!!warning macOS Docker memory allocation
+!!! warning macOS Docker memory allocation
 On macOS, the default amount of memory available for Docker might not be enough to get Airflow up and running. Allocate at least 4GB of memory for the Docker Engine (ideally 8GB). [More info](https://airflow.apache.org/docs/apache-airflow/stable/howto/docker-compose/index.html#)
 
 ## Configure Airflow access to Google Cloud Platform
 
-!!!warning Specifying Google Cloud parameters
+!!! warning Specifying Google Cloud parameters
 Run the next two command with the appropriate Google Cloud project ID and service account name to ensure the correct Google default application credentials are set up.
 
 Authenticate to Google Cloud:
@@ -37,7 +37,7 @@ cd src/airflow
 
 ### Build Docker image
 
-!!!note Custom Docker image for Airflow
+!!! note Custom Docker image for Airflow
 The custom Dockerfile built by the command below extends the official [Airflow Docker Compose YAML](https://airflow.apache.org/docs/apache-airflow/stable/docker-compose.yaml). We add support for Google Cloud SDK, Google Dataproc operators, and access to GCP credentials.
 
 ```bash
@@ -46,7 +46,7 @@ docker build . --tag extending_airflow:latest
 
 ### Set Airflow user ID
 
-!!!note Setting Airflow user ID
+!!! note Setting Airflow user ID
 These commands allow Airflow running inside Docker to access the credentials file which was generated earlier.
 
 ```bash
diff --git a/docs/howto/.pages b/docs/howto/.pages
new file mode 100644
index 000000000..eb8058f0f
--- /dev/null
+++ b/docs/howto/.pages
@@ -0,0 +1 @@
+title: How-to
diff --git a/docs/howto/_howto.md b/docs/howto/_howto.md
new file mode 100644
index 000000000..1e8e26b83
--- /dev/null
+++ b/docs/howto/_howto.md
@@ -0,0 +1,5 @@
+# How-to
+
+This page contains a collection of how-to guides for the project.
+
+For additional information please visit [https://community.opentargets.org/](https://community.opentargets.org/)
diff --git a/docs/howto/run_step_in_cli.md b/docs/howto/run_step_in_cli.md
new file mode 100644
index 000000000..925a17f28
--- /dev/null
+++ b/docs/howto/run_step_in_cli.md
@@ -0,0 +1,44 @@
+---
+Title: Run step in CLI
+---
+
+# Run step in CLI
+
+To run a step in the command line interface (CLI), you need to know the step's name. To list what steps are avaiable in your current environment, simply run `otg` with no arguments. This will list all the steps:
+
+```
+You must specify 'step', e.g, step=<OPTION>
+Available options:
+        clump
+        colocalisation
+        eqtl_catalogue
+        finngen_studies
+        finngen_sumstat_preprocess
+        gene_index
+        gwas_catalog_ingestion
+        gwas_catalog_sumstat_preprocess
+        ld_index
+        locus_to_gene
+        overlaps
+        pics
+        ukbiobank
+        variant_annotation
+        variant_index
+        variant_to_gene
+
+Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.
+```
+
+As indicated, you can run a step by specifying the step's name with the `step` argument. For example, to run the `gene_index` step, you can run:
+
+```bash
+otg step=gene_index
+```
+
+In most occassions, some mandatory values will be required to run the step. For example, the `gene_index` step requires the `step.target_path` and `step.gene_index_path` argument to be specified. You can complete the necessary arguments by adding them to the command line:
+
+```bash
+otg step=gene_index step.target_path=/path/to/target step.gene_index_path=/path/to/gene_index
+```
+
+You can find more about the available steps in the [documentation](../python_api/step/_step.md).
diff --git a/docs/howto/run_step_using_config.md b/docs/howto/run_step_using_config.md
new file mode 100644
index 000000000..68f302b50
--- /dev/null
+++ b/docs/howto/run_step_using_config.md
@@ -0,0 +1,49 @@
+---
+Title: Run step using config
+---
+
+# Run step using YAML config
+
+It's possible to parametrise the functionality of a step using a YAML configuration file. This is useful when you want to run a step multiple times with different parameters or simply to avoid having to specify the same parameters every time you run a step.
+
+!!! info Configuration files using Hydra
+The package uses [Hydra](https://hydra.cc) to handle configuration files. For more information, please visit the [Hydra documentation](https://hydra.cc/docs/intro/).
+
+To run a step using a configuration file, you need to create a configuration file in YAML format.
+
+```{ .sh .no-copy }
+config/
+├─ step/
+│  └─ my_gene_index.md
+└─ my_config.yml
+```
+
+The configuration file should contain the parameters you want to use to run the step. For example, to run the `gene_index` step, you need to specify the `step.target_path` and `step.gene_index_path` parameters. The configuration file should look like this:
+
+=== "my_config.yaml"
+
+    ``` yaml
+    defaults:
+        - config
+        - _self_
+    ```
+
+    This config file will specify that your configuration file will inherit the default configuration (`config`) and everything provided (`_self_`) will overwrite the default configuration.
+
+=== "step/my_gene_index.md"
+
+    ``` yaml
+    defaults:
+        - gene_index
+
+    target_path: /path/to/target
+    gene_index_path: /path/to/gene_index
+    ```
+
+    This config file will inherit the default configuration for the `gene_index` step and overwrite the `target_path` and `gene_index_path` parameters.
+
+Once you have created the configuration file, you can run your own new `my_gene_index`:
+
+```bash
+otg step=my_gene_index --config-dir=config --config-name=my_config
+```
diff --git a/docs/index.md b/docs/index.md
index 006477ec3..791779028 100644
--- a/docs/index.md
+++ b/docs/index.md
@@ -29,4 +29,4 @@ Ingestion and analysis of genetic and functional genomic data for the identifica
 
 This project is still in experimental phase. Please refer to the [roadmap section](roadmap.md) for more information.
 
-For all development information, including running the code, troubleshooting, or contributing, see the [development section](./development/).
+For all development information, including running the code, troubleshooting, or contributing, see the [development section](development/_development.md).
diff --git a/docs/python_api/step/gwas_catalog_inclusion.md b/docs/python_api/step/gwas_catalog_inclusion.md
index b9f088a76..c5a6b89c7 100644
--- a/docs/python_api/step/gwas_catalog_inclusion.md
+++ b/docs/python_api/step/gwas_catalog_inclusion.md
@@ -2,4 +2,4 @@
 title: Generate inclusion and exclusions lists for GWAS Catalog study ingestion.
 ---
 
-::: otg.gwas_catalog_study_inclusion.GWASCatalogInclusionGenerator
+::: otg.gwas_catalog_study_inclusion.GWASCatalogStudyInclusionGenerator
diff --git a/docs/usage.md b/docs/usage.md
deleted file mode 100644
index 4d6c97d1b..000000000
--- a/docs/usage.md
+++ /dev/null
@@ -1,10 +0,0 @@
----
-Title: How-to
-hide:
-  - navigation
-  - toc
----
-
-# How-to
-
-TBC
diff --git a/mkdocs.yml b/mkdocs.yml
index e15a1af8d..2426f70fc 100644
--- a/mkdocs.yml
+++ b/mkdocs.yml
@@ -1,19 +1,22 @@
 site_name: Open Targets Genetics
 
 nav:
-  - installation.md
-  - usage.md
+  - Home: index.md
+  - Installation: installation.md
+  - ... | howto/**
+  - Roadmap: roadmap.md
   - ... | development/**
   - ... | python_api/**
 
 plugins:
   - search
-  - awesome-pages
+  - awesome-pages:
+      collapse_single_pages: true
   - mkdocstrings:
       handlers:
         python:
           options:
-            filters: ["!^_", "!__new__"]
+            filters: ["!^_", "!__new__", "__init__"]
             show_signature_annotations: true
             show_root_heading: true
   - section-index
@@ -41,6 +44,9 @@ markdown_extensions:
   - pymdownx.superfences
   - toc:
       permalink: true
+  - pymdownx.tabbed:
+      alternate_style: true
+      combine_header_slug: true
 
 hooks:
   - src/utils/schemadocs.py
diff --git a/src/airflow/dags/common_airflow.py b/src/airflow/dags/common_airflow.py
index b2deb9268..900036d6d 100644
--- a/src/airflow/dags/common_airflow.py
+++ b/src/airflow/dags/common_airflow.py
@@ -40,7 +40,7 @@
 
 # CLI configuration.
 CLUSTER_CONFIG_DIR = "/config"
-CONFIG_NAME = "config"
+CONFIG_NAME = "ot_config"
 PYTHON_CLI = "cli.py"
 
 # Shared DAG construction parameters.
diff --git a/src/airflow/dags/configs/dag.yaml b/src/airflow/dags/configs/dag.yaml
index caf55b373..e1c9ea627 100644
--- a/src/airflow/dags/configs/dag.yaml
+++ b/src/airflow/dags/configs/dag.yaml
@@ -1,18 +1,18 @@
-- id: "gene_index"
-- id: "gwas_catalog"
-- id: "variant_index"
+- id: "ot_gene_index"
+- id: "ot_gwas_catalog"
+- id: "ot_variant_index"
   prerequisites:
-    - "gwas_catalog"
-- id: "v2g"
+    - "ot_gwas_catalog"
+- id: "ot_variant_to_gene"
   prerequisites:
-    - "variant_index"
-    - "gene_index"
-- id: "study_locus_overlap"
+    - "ot_variant_index"
+    - "ot_gene_index"
+- id: "ot_study_locus_overlap"
   prerequisites:
-    - "gwas_catalog"
-- id: "locus_to_gene"
+    - "ot_gwas_catalog"
+- id: "ot_locus_to_gene"
   prerequisites:
-    - "gwas_catalog"
-    - "variant_index"
-    - "v2g"
-    - "study_locus_overlap"
+    - "ot_gwas_catalog"
+    - "ot_variant_index"
+    - "ot_variant_to_gene"
+    - "ot_study_locus_overlap"
diff --git a/src/airflow/dags/dag_preprocess.py b/src/airflow/dags/dag_preprocess.py
index f0dde085e..728d5932f 100644
--- a/src/airflow/dags/dag_preprocess.py
+++ b/src/airflow/dags/dag_preprocess.py
@@ -9,11 +9,10 @@
 CLUSTER_NAME = "otg-preprocess"
 
 ALL_STEPS = [
-    "finngen",
-    "eqtl_catalogue",
-    "ld_index",
-    "variant_annotation",
-    "ukbiobank",
+    "ot_eqtl_catalogue",
+    "ot_ld_index",
+    "ot_variant_annotation",
+    "ot_ukbiobank",
 ]
 
 
diff --git a/src/airflow/dags/finngen_harmonisation.py b/src/airflow/dags/finngen_harmonisation.py
index ad88e695c..e2f2d1b4a 100644
--- a/src/airflow/dags/finngen_harmonisation.py
+++ b/src/airflow/dags/finngen_harmonisation.py
@@ -53,7 +53,7 @@ def submit_jobs(**kwargs: Any) -> None:
             print("Submitting job for study: ", study_id)  # noqa: T201
             common.submit_pyspark_job_no_operator(
                 cluster_name=CLUSTER_NAME,
-                step_id="finngen_sumstat_preprocess",
+                step_id="ot_finngen_sumstat_preprocess",
                 other_args=[
                     f"step.raw_sumstats_path=gs://{SUMMARY_STATS_BUCKET_NAME}/{input_path}",
                     f"step.out_sumstats_path={SUMSTATS_PARQUET}/{study_id}.parquet",
diff --git a/src/airflow/dags/finngen_preprocess.py b/src/airflow/dags/finngen_preprocess.py
index e40248cd6..32dc68c9c 100644
--- a/src/airflow/dags/finngen_preprocess.py
+++ b/src/airflow/dags/finngen_preprocess.py
@@ -26,7 +26,7 @@
 ):
     study_index = common.submit_step(
         cluster_name=CLUSTER_NAME,
-        step_id="finngen_studies",
+        step_id="ot_finngen_studies",
         task_id="finngen_studies",
     )
 
@@ -54,7 +54,7 @@
 
     pics = common.submit_step(
         cluster_name=CLUSTER_NAME,
-        step_id="pics",
+        step_id="ot_pics",
         task_id="finngen_pics",
         other_args=[
             f"step.study_locus_ld_annotated_in={LD_CLUMPED}",
diff --git a/src/airflow/dags/gwas_catalog_harmonisation.py b/src/airflow/dags/gwas_catalog_harmonisation.py
index 41815caa2..7e7790224 100644
--- a/src/airflow/dags/gwas_catalog_harmonisation.py
+++ b/src/airflow/dags/gwas_catalog_harmonisation.py
@@ -92,7 +92,7 @@ def submit_jobs(**kwargs: Any) -> None:
             print("Submitting job for study: ", study_id)  # noqa: T201
             common.submit_pyspark_job_no_operator(
                 cluster_name=CLUSTER_NAME,
-                step_id="gwas_catalog_sumstat_preprocess",
+                step_id="ot_gwas_catalog_sumstat_preprocess",
                 other_args=[
                     f"step.raw_sumstats_path=gs://{SUMMARY_STATS_BUCKET_NAME}/{input_path}",
                     f"step.out_sumstats_path=gs://{SUMMARY_STATS_BUCKET_NAME}/harmonised/{study_id}.parquet",
diff --git a/src/airflow/dags/gwas_catalog_preprocess.py b/src/airflow/dags/gwas_catalog_preprocess.py
index c3d094dd2..b7dd1b08e 100644
--- a/src/airflow/dags/gwas_catalog_preprocess.py
+++ b/src/airflow/dags/gwas_catalog_preprocess.py
@@ -69,7 +69,7 @@ def upload_harmonized_study_list(
         # Generate inclusion list:
         curation_calculate_inclusion_list = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="gwas_study_inclusion",
+            step_id="ot_gwas_catalog_study_inclusion",
             task_id="catalog_curation_inclusion_list",
             other_args=[
                 "step.criteria=curation",
@@ -82,7 +82,7 @@ def upload_harmonized_study_list(
         # Ingest curated associations from GWAS Catalog:
         curation_ingest_data = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="gwas_catalog_ingestion",
+            step_id="ot_gwas_catalog_ingestion",
             task_id="ingest_curated_gwas_catalog_data",
             other_args=[f"step.inclusion_list_path={MANIFESTS_PATH}manifest_curation"],
         )
@@ -90,7 +90,7 @@ def upload_harmonized_study_list(
         # Run LD-annotation and clumping on curated data:
         curation_ld_clumping = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="ld_based_clumping",
+            step_id="ot_ld_based_clumping",
             task_id="catalog_curation_ld_clumping",
             other_args=[
                 f"step.study_locus_input_path={RELEASEBUCKET}/study_locus/catalog_curated",
@@ -103,7 +103,7 @@ def upload_harmonized_study_list(
         # Do PICS based finemapping:
         curation_pics = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="pics",
+            step_id="ot_pics",
             task_id="catalog_curation_pics",
             other_args=[
                 f"step.study_locus_ld_annotated_in={RELEASEBUCKET}/study_locus/ld_clumped/catalog_curated",
@@ -126,7 +126,7 @@ def upload_harmonized_study_list(
         # Generate inclusion study lists:
         summary_stats_calculate_inclusion_list = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="gwas_study_inclusion",
+            step_id="ot_gwas_catalog_study_inclusion",
             task_id="catalog_sumstats_inclusion_list",
             other_args=[
                 "step.criteria=summary_stats",
@@ -139,7 +139,7 @@ def upload_harmonized_study_list(
         # Run window-based clumping:
         summary_stats_window_based_clumping = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="window_based_clumping",
+            step_id="ot_window_based_clumping",
             task_id="catalog_sumstats_window_clumping",
             other_args=[
                 f"step.summary_statistics_input_path={SUMSTATS}",
@@ -151,7 +151,7 @@ def upload_harmonized_study_list(
         # Run LD based clumping:
         summary_stats_ld_clumping = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="ld_based_clumping",
+            step_id="ot_ld_based_clumping",
             task_id="catalog_sumstats_ld_clumping",
             other_args=[
                 f"step.study_locus_input_path={RELEASEBUCKET}/study_locus/window_clumped/from_sumstats/catalog",
@@ -164,7 +164,7 @@ def upload_harmonized_study_list(
         # Run PICS finemapping:
         summary_stats_pics = common.submit_step(
             cluster_name=CLUSTER_NAME,
-            step_id="pics",
+            step_id="ot_pics",
             task_id="catalog_sumstats_pics",
             other_args=[
                 f"step.study_locus_ld_annotated_in={RELEASEBUCKET}/study_locus/ld_clumped/from_sumstats/catalog",
diff --git a/src/otg/cli.py b/src/otg/cli.py
index e30d587a6..088336877 100644
--- a/src/otg/cli.py
+++ b/src/otg/cli.py
@@ -3,15 +3,19 @@
 
 import hydra
 from hydra.utils import instantiate
-from omegaconf import DictConfig, OmegaConf
+from omegaconf import OmegaConf
+
+from otg.config import Config, register_config
+
+register_config()
 
 
 @hydra.main(version_base="1.3", config_path=None, config_name="config")
-def main(cfg: DictConfig) -> None:
+def main(cfg: Config) -> None:
     """OTG ETL CLI.
 
     Args:
-        cfg (DictConfig): hydra configuration object
+        cfg (Config): configuration object.
     """
     print(OmegaConf.to_yaml(cfg))  # noqa: T201
     # Initialise and run step
diff --git a/src/otg/colocalisation.py b/src/otg/colocalisation.py
index 150304485..a65501de1 100644
--- a/src/otg/colocalisation.py
+++ b/src/otg/colocalisation.py
@@ -1,41 +1,39 @@
 """Step to generate colocalisation results."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.dataset.study_index import StudyIndex
 from otg.dataset.study_locus import CredibleInterval, StudyLocus
 from otg.method.colocalisation import ECaviar
 
 
-@dataclass
 class ColocalisationStep:
     """Colocalisation step.
 
-    This workflow runs colocalization analyses that assess the degree to which independent signals of the association share the same causal variant in a region of the genome, typically limited by linkage disequilibrium (LD).
-
-    Attributes:
-        session (Session): Session object.
-        credible_set_path (DictConfig): Input credible sets path.
-        coloc_path (DictConfig): Output Colocalisation path.
+    This workflow runs colocalisation analyses that assess the degree to which independent signals of the association share the same causal variant in a region of the genome, typically limited by linkage disequilibrium (LD).
     """
 
-    session: Session = MISSING
-    credible_set_path: str = MISSING
-    study_index_path: str = MISSING
-    coloc_path: str = MISSING
-
-    def __post_init__(self: ColocalisationStep) -> None:
-        """Run step."""
+    def __init__(
+        self,
+        session: Session,
+        credible_set_path: str,
+        study_index_path: str,
+        coloc_path: str,
+    ) -> None:
+        """Run Colocalisation step.
+
+        Args:
+            session (Session): Session object.
+            credible_set_path (str): Input credible sets path.
+            study_index_path (str): Input study index path.
+            coloc_path (str): Output Colocalisation path.
+        """
         # Extract
         credible_set = StudyLocus.from_parquet(
-            self.session, self.credible_set_path, recursiveFileLookup=True
+            session, credible_set_path, recursiveFileLookup=True
         )
         si = StudyIndex.from_parquet(
-            self.session, self.study_index_path, recursiveFileLookup=True
+            session, study_index_path, recursiveFileLookup=True
         )
 
         # Transform
@@ -45,4 +43,4 @@ def __post_init__(self: ColocalisationStep) -> None:
         ecaviar_results = ECaviar.colocalise(overlaps)
 
         # Load
-        ecaviar_results.df.write.mode(self.session.write_mode).parquet(self.coloc_path)
+        ecaviar_results.df.write.mode(session.write_mode).parquet(coloc_path)
diff --git a/src/otg/config.py b/src/otg/config.py
new file mode 100644
index 000000000..1506c134e
--- /dev/null
+++ b/src/otg/config.py
@@ -0,0 +1,375 @@
+"""Interface for application configuration."""
+import os
+from dataclasses import dataclass, field
+from typing import Any, Dict, List
+
+from hail import __file__ as hail_location
+from hydra.core.config_store import ConfigStore
+from omegaconf import MISSING
+
+
+@dataclass
+class SessionConfig:
+    """Session configuration."""
+
+    start_hail: bool = False
+    write_mode: str = "errorifexists"
+    spark_uri: str = "local[*]"
+    hail_home: str = os.path.dirname(hail_location)
+    extended_spark_conf: dict[str, str] | None = None
+    _target_: str = "otg.common.session.Session"
+
+
+@dataclass
+class StepConfig:
+    """Base step configuration."""
+
+    session: SessionConfig
+    defaults: List[Any] = field(
+        default_factory=lambda: [{"session": "base_session"}, "_self_"]
+    )
+
+
+@dataclass
+class ColocalisationConfig(StepConfig):
+    """Colocalisation step configuration."""
+
+    study_locus_path: str = MISSING
+    study_index_path: str = MISSING
+    coloc_path: str = MISSING
+    priorc1: float = 1e-4
+    priorc2: float = 1e-4
+    priorc12: float = 1e-5
+    _target_: str = "otg.colocalisation.ColocalisationStep"
+
+
+@dataclass
+class GeneIndexConfig(StepConfig):
+    """Gene index step configuration."""
+
+    target_path: str = MISSING
+    gene_index_path: str = MISSING
+    _target_: str = "otg.gene_index.GeneIndexStep"
+
+
+@dataclass
+class GWASCatalogStudyCurationConfig(StepConfig):
+    """GWAS Catalog study curation step configuration."""
+
+    catalog_study_files: list[str] = MISSING
+    catalog_ancestry_files: list[str] = MISSING
+    catalog_sumstats_lut: str = MISSING
+    gwas_catalog_study_curation_out: str = MISSING
+    gwas_catalog_study_curation_file: str = MISSING
+    _target_: str = "otg.gwas_catalog_study_curation.GWASCatalogStudyCurationStep"
+
+
+@dataclass
+class GWASCatalogStudyInclusionConfig(StepConfig):
+    """GWAS Catalog study inclusion step configuration."""
+
+    catalog_study_files: list[str] = MISSING
+    catalog_ancestry_files: list[str] = MISSING
+    catalog_associations_file: str = MISSING
+    gwas_catalog_study_curation_file: str = MISSING
+    variant_annotation_path: str = MISSING
+    harmonised_study_file: str = MISSING
+    criteria: str = MISSING
+    inclusion_list_path: str = MISSING
+    exclusion_list_path: str = MISSING
+    _target_: str = "otg.gwas_catalog_study_inclusion.GWASCatalogStudyInclusionStep"
+
+
+@dataclass
+class GWASCatalogIngestionConfig(StepConfig):
+    """GWAS Catalog ingestion step configuration."""
+
+    catalog_study_files: list[str] = MISSING
+    catalog_ancestry_files: list[str] = MISSING
+    catalog_sumstats_lut: str = MISSING
+    catalog_associations_file: str = MISSING
+    variant_annotation_path: str = MISSING
+    catalog_studies_out: str = MISSING
+    catalog_associations_out: str = MISSING
+    gwas_catalog_study_curation_file: str | None = None
+    inclusion_list_path: str | None = None
+    _target_: str = "otg.gwas_catalog_ingestion.GWASCatalogIngestionStep"
+
+
+@dataclass
+class GWASCatalogSumstatsPreprocessConfig(StepConfig):
+    """GWAS Catalog sumstat preprocess step configuration."""
+
+    raw_sumstats_path: str = MISSING
+    out_sumstats_path: str = MISSING
+    _target_: str = (
+        "otg.gwas_catalog_sumstat_preprocess.GWASCatalogSumstatsPreprocessStep"
+    )
+
+
+@dataclass
+class EqtlCatalogueConfig(StepConfig):
+    """eQTL Catalogue step configuration."""
+
+    eqtl_catalogue_paths_imported: str = MISSING
+    eqtl_catalogue_study_index_out: str = MISSING
+    eqtl_catalogue_summary_stats_out: str = MISSING
+    _target_: str = "otg.eqtl_catalogue.EqtlCatalogueStep"
+
+
+@dataclass
+class FinngenStudiesConfig(StepConfig):
+    """FinnGen study index step configuration."""
+
+    finngen_study_index_out: str = MISSING
+    _target_: str = "otg.finngen_studies.FinnGenStudiesStep"
+
+
+@dataclass
+class FinngenSumstatPreprocessConfig(StepConfig):
+    """FinnGen study index step configuration."""
+
+    raw_sumstats_path: str = MISSING
+    out_sumstats_path: str = MISSING
+    _target_: str = "otg.finngen_sumstat_preprocess.FinnGenSumstatPreprocessStep"
+
+
+@dataclass
+class LDIndexConfig(StepConfig):
+    """LD index step configuration."""
+
+    session: Any = field(
+        default_factory=lambda: {
+            "start_hail": True,
+        }
+    )
+    min_r2: float = 0.5
+    ld_index_out: str = MISSING
+    _target_: str = "otg.ld_index.LDIndexStep"
+
+
+@dataclass
+class LDBasedClumpingConfig(StepConfig):
+    """LD based clumping step configuration."""
+
+    study_locus_input_path: str = MISSING
+    study_index_path: str = MISSING
+    ld_index_path: str = MISSING
+    clumped_study_locus_output_path: str = MISSING
+    _target_: str = "otg.ld_based_clumping.LDBasedClumpingStep"
+
+
+@dataclass
+class LocusToGeneConfig(StepConfig):
+    """Locus to gene step configuration."""
+
+    session: Any = field(
+        default_factory=lambda: {
+            "extended_spark_conf": {"spark.dynamicAllocation.enabled": "false"}
+        }
+    )
+    run_mode: str = MISSING
+    model_path: str = MISSING
+    predictions_path: str = MISSING
+    credible_set_path: str = MISSING
+    variant_gene_path: str = MISSING
+    colocalisation_path: str = MISSING
+    study_index_path: str = MISSING
+    study_locus_overlap_path: str = MISSING
+    gold_standard_curation_path: str = MISSING
+    gene_interactions_path: str = MISSING
+    features_list: list[str] = field(
+        default_factory=lambda: [
+            # average distance of all tagging variants to gene TSS
+            "distanceTssMean",
+            # # minimum distance of all tagging variants to gene TSS
+            "distanceTssMinimum",
+            # # maximum vep consequence score of the locus 95% credible set among all genes in the vicinity
+            "vepMaximumNeighborhood",
+            # # maximum vep consequence score of the locus 95% credible set split by gene
+            "vepMaximum",
+            # max clpp for each (study, locus, gene) aggregating over all eQTLs
+            "eqtlColocClppMaximum",
+            # max clpp for each (study, locus) aggregating over all eQTLs
+            "eqtlColocClppMaximumNeighborhood",
+            # max clpp for each (study, locus, gene) aggregating over all pQTLs
+            # "pqtlColocClppMaximum",
+            # max clpp for each (study, locus) aggregating over all pQTLs
+            # "pqtlColocClppMaximumNeighborhood",
+            # max clpp for each (study, locus, gene) aggregating over all sQTLs
+            # "sqtlColocClppMaximum",
+            # max clpp for each (study, locus) aggregating over all sQTLs
+            # "sqtlColocClppMaximumNeighborhood",
+            # # max log-likelihood ratio value for each (study, locus, gene) aggregating over all eQTLs
+            # "eqtlColocLlrLocalMaximum",
+            # # max log-likelihood ratio value for each (study, locus) aggregating over all eQTLs
+            # "eqtlColocLlpMaximumNeighborhood",
+            # # max log-likelihood ratio value for each (study, locus, gene) aggregating over all pQTLs
+            # "pqtlColocLlrLocalMaximum",
+            # # max log-likelihood ratio value for each (study, locus) aggregating over all pQTLs
+            # "pqtlColocLlpMaximumNeighborhood",
+            # # max log-likelihood ratio value for each (study, locus, gene) aggregating over all sQTLs
+            # "sqtlColocLlrLocalMaximum",
+            # # max log-likelihood ratio value for each (study, locus) aggregating over all sQTLs
+            # "sqtlColocLlpMaximumNeighborhood",
+        ]
+    )
+    hyperparameters: dict[str, Any] = field(
+        default_factory=lambda: {
+            "max_depth": 5,
+            "loss_function": "binary:logistic",
+        }
+    )
+    wandb_run_name: str | None = None
+    perform_cross_validation: bool = False
+    _target_: str = "otg.l2g.LocusToGeneStep"
+
+
+@dataclass
+class OverlapsIndexConfig(StepConfig):
+    """Overlaps step configuration."""
+
+    study_locus_path: str = MISSING
+    study_index_path: str = MISSING
+    overlaps_index_out: str = MISSING
+    _target_: str = "otg.overlaps.OverlapsIndexStep"
+
+
+@dataclass
+class PICSConfig(StepConfig):
+    """PICS step configuration."""
+
+    study_locus_ld_annotated_in: str = MISSING
+    picsed_study_locus_out: str = MISSING
+    _target_: str = "otg.pics.PICSStep"
+
+
+@dataclass
+class UKBiobankConfig(StepConfig):
+    """UKBiobank step configuration."""
+
+    ukbiobank_manifest: str = MISSING
+    ukbiobank_study_index_out: str = MISSING
+    _target_: str = "otg.ukbiobank.UKBiobankStep"
+
+
+@dataclass
+class VariantAnnotationConfig(StepConfig):
+    """Variant annotation step configuration."""
+
+    session: Any = field(
+        default_factory=lambda: {
+            "start_hail": True,
+        }
+    )
+    variant_annotation_path: str = MISSING
+    _target_: str = "otg.variant_annotation.VariantAnnotationStep"
+
+
+@dataclass
+class VariantIndexConfig(StepConfig):
+    """Variant index step configuration."""
+
+    variant_annotation_path: str = MISSING
+    credible_set_path: str = MISSING
+    variant_index_path: str = MISSING
+    _target_: str = "otg.variant_index.VariantIndexStep"
+
+
+@dataclass
+class V2GConfig(StepConfig):
+    """V2G step configuration."""
+
+    variant_index_path: str = MISSING
+    variant_annotation_path: str = MISSING
+    gene_index_path: str = MISSING
+    vep_consequences_path: str = MISSING
+    liftover_chain_file_path: str = MISSING
+    liftover_max_length_difference: int = 100
+    max_distance: int = 500_000
+    approved_biotypes: List[str] = field(
+        default_factory=lambda: [
+            "protein_coding",
+            "3prime_overlapping_ncRNA",
+            "antisense",
+            "bidirectional_promoter_lncRNA",
+            "IG_C_gene",
+            "IG_D_gene",
+            "IG_J_gene",
+            "IG_V_gene",
+            "lincRNA",
+            "macro_lncRNA",
+            "non_coding",
+            "sense_intronic",
+            "sense_overlapping",
+        ]
+    )
+    interval_sources: Dict[str, str] = field(default_factory=dict)
+    v2g_path: str = MISSING
+    _target_: str = "otg.v2g.V2GStep"
+
+
+@dataclass
+class WindowBasedClumpingStep(StepConfig):
+    """Window-based clumping step configuration."""
+
+    summary_statistics_input_path: str = MISSING
+    study_locus_output_path: str = MISSING
+    inclusion_list_path: str = MISSING
+    locus_collect_distance: str | None = None
+
+    _target_: str = "otg.clump.WindowBasedClumpingStep"
+
+
+@dataclass
+class Config:
+    """Application configuration."""
+
+    # this is unfortunately verbose due to @dataclass limitations
+    defaults: List[Any] = field(default_factory=lambda: ["_self_", {"step": MISSING}])
+    step: StepConfig = MISSING
+    datasets: dict[str, str] = field(default_factory=dict)
+
+
+def register_config() -> None:
+    """Register configuration."""
+    cs = ConfigStore.instance()
+    cs.store(name="config", node=Config)
+    cs.store(group="step/session", name="base_session", node=SessionConfig)
+    cs.store(group="step", name="colocalisation", node=ColocalisationConfig)
+    cs.store(group="step", name="eqtl_catalogue", node=EqtlCatalogueConfig)
+    cs.store(group="step", name="gene_index", node=GeneIndexConfig)
+    cs.store(
+        group="step",
+        name="gwas_catalog_study_curation",
+        node=GWASCatalogStudyCurationConfig,
+    )
+    cs.store(
+        group="step",
+        name="gwas_catalog_study_inclusion",
+        node=GWASCatalogStudyInclusionConfig,
+    )
+    cs.store(
+        group="step", name="gwas_catalog_ingestion", node=GWASCatalogIngestionConfig
+    )
+    cs.store(
+        group="step",
+        name="gwas_catalog_sumstat_preprocess",
+        node=GWASCatalogSumstatsPreprocessConfig,
+    )
+    cs.store(group="step", name="ld_based_clumping", node=LDBasedClumpingConfig)
+    cs.store(group="step", name="ld_index", node=LDIndexConfig)
+    cs.store(group="step", name="locus_to_gene", node=LocusToGeneConfig)
+    cs.store(group="step", name="finngen_studies", node=FinngenStudiesConfig)
+    cs.store(
+        group="step",
+        name="finngen_sumstat_preprocess",
+        node=FinngenSumstatPreprocessConfig,
+    )
+    cs.store(group="step", name="overlaps", node=OverlapsIndexConfig)
+    cs.store(group="step", name="pics", node=PICSConfig)
+    cs.store(group="step", name="ukbiobank", node=UKBiobankConfig)
+    cs.store(group="step", name="variant_annotation", node=VariantAnnotationConfig)
+    cs.store(group="step", name="variant_index", node=VariantIndexConfig)
+    cs.store(group="step", name="variant_to_gene", node=V2GConfig)
+    cs.store(group="step", name="window_based_clumping", node=WindowBasedClumpingStep)
diff --git a/src/otg/eqtl_catalogue.py b/src/otg/eqtl_catalogue.py
index 362701878..9898a8e29 100644
--- a/src/otg/eqtl_catalogue.py
+++ b/src/otg/eqtl_catalogue.py
@@ -2,37 +2,32 @@
 
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.datasource.eqtl_catalogue.study_index import EqtlCatalogueStudyIndex
 from otg.datasource.eqtl_catalogue.summary_stats import EqtlCatalogueSummaryStats
 
 
-@dataclass
 class EqtlCatalogueStep:
-    """eQTL Catalogue ingestion step.
-
-    Attributes:
-        session (Session): Session object.
-        eqtl_catalogue_paths_imported (str): eQTL Catalogue input files for the harmonised and imported data.
-        eqtl_catalogue_study_index_out (str): Output path for the eQTL Catalogue study index dataset.
-        eqtl_catalogue_summary_stats_out (str): Output path for the eQTL Catalogue summary stats.
-    """
-
-    session: Session = MISSING
-
-    eqtl_catalogue_paths_imported: str = MISSING
-    eqtl_catalogue_study_index_out: str = MISSING
-    eqtl_catalogue_summary_stats_out: str = MISSING
-
-    def __post_init__(self: EqtlCatalogueStep) -> None:
-        """Run step."""
+    """eQTL Catalogue ingestion step."""
+
+    def __init__(
+        self,
+        session: Session,
+        eqtl_catalogue_paths_imported: str,
+        eqtl_catalogue_study_index_out: str,
+        eqtl_catalogue_summary_stats_out: str,
+    ) -> None:
+        """Run eQTL Catalogue ingestion step.
+
+        Args:
+            session (Session): Session object.
+            eqtl_catalogue_paths_imported (str): Input eQTL Catalogue study index path.
+            eqtl_catalogue_study_index_out (str): Output eQTL Catalogue study index path.
+            eqtl_catalogue_summary_stats_out (str): Output eQTL Catalogue summary stats path.
+        """
         # Fetch study index.
-        df = self.session.spark.read.option("delimiter", "\t").csv(
-            self.eqtl_catalogue_paths_imported, header=True
+        df = session.spark.read.option("delimiter", "\t").csv(
+            eqtl_catalogue_paths_imported, header=True
         )
         # Process partial study index.  At this point, it is not complete because we don't have the gene IDs, which we
         # will only get once the summary stats are ingested.
@@ -41,7 +36,7 @@ def __post_init__(self: EqtlCatalogueStep) -> None:
         # Fetch summary stats.
         input_filenames = [row.summarystatsLocation for row in study_index_df.collect()]
         summary_stats_df = (
-            self.session.spark.read.option("delimiter", "\t")
+            session.spark.read.option("delimiter", "\t")
             .csv(input_filenames, header=True)
             .repartition(1280)
         )
@@ -55,13 +50,13 @@ def __post_init__(self: EqtlCatalogueStep) -> None:
         ).df
 
         # Write study index.
-        study_index_df.write.mode(self.session.write_mode).parquet(
-            self.eqtl_catalogue_study_index_out
+        study_index_df.write.mode(session.write_mode).parquet(
+            eqtl_catalogue_study_index_out
         )
         # Write summary stats.
         (
             summary_stats_df.sortWithinPartitions("position")
             .write.partitionBy("chromosome")
-            .mode(self.session.write_mode)
-            .parquet(self.eqtl_catalogue_summary_stats_out)
+            .mode(session.write_mode)
+            .parquet(eqtl_catalogue_summary_stats_out)
         )
diff --git a/src/otg/finngen_studies.py b/src/otg/finngen_studies.py
index 9a1d800e8..e7d67f113 100644
--- a/src/otg/finngen_studies.py
+++ b/src/otg/finngen_studies.py
@@ -1,31 +1,22 @@
-"""Step to run FinnGen study table ingestion."""
+"""Step to run FinnGen study index generation."""
 
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.datasource.finngen.study_index import FinnGenStudyIndex
 
 
-@dataclass
 class FinnGenStudiesStep:
-    """FinnGen study index generation step.
-
-    Attributes:
-        session (Session): Session object.
-        finngen_study_index_out (str): Output path for the FinnGen study index dataset.
-    """
+    """FinnGen study index generation step."""
 
-    session: Session = MISSING
-    finngen_study_index_out: str = MISSING
-    finngen_summary_stats_out: str = MISSING
+    def __init__(self, session: Session, finngen_study_index_out: str) -> None:
+        """Run FinnGen study index generation step.
 
-    def __post_init__(self: FinnGenStudiesStep) -> None:
-        """Run step."""
+        Args:
+            session (Session): Session object.
+            finngen_study_index_out (str): Output FinnGen study index path.
+        """
         # Fetch study index.
-        FinnGenStudyIndex.from_source(self.session.spark).df.write.mode(
-            self.session.write_mode
-        ).parquet(self.finngen_study_index_out)
+        FinnGenStudyIndex.from_source(session.spark).df.write.mode(
+            session.write_mode
+        ).parquet(finngen_study_index_out)
diff --git a/src/otg/finngen_sumstat_preprocess.py b/src/otg/finngen_sumstat_preprocess.py
index 959c000dc..802c9aa77 100644
--- a/src/otg/finngen_sumstat_preprocess.py
+++ b/src/otg/finngen_sumstat_preprocess.py
@@ -2,35 +2,26 @@
 
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.datasource.finngen.summary_stats import FinnGenSummaryStats
 
 
-@dataclass
 class FinnGenSumstatPreprocessStep:
-    """FinnGen sumstats preprocessing.
-
-    Attributes:
-        session (Session): Session object.
-        finngen_study_index_out (str): Output path for the FinnGen study index dataset.
-        finngen_summary_stats_out (str): Output path for the FinnGen summary statistics.
-    """
-
-    session: Session = MISSING
-    raw_sumstats_path: str = MISSING
-    out_sumstats_path: str = MISSING
-
-    def __post_init__(self: FinnGenSumstatPreprocessStep) -> None:
-        """Run step."""
+    """FinnGen sumstats preprocessing."""
+
+    def __init__(
+        self, session: Session, raw_sumstats_path: str, out_sumstats_path: str
+    ) -> None:
+        """Run FinnGen summary stats preprocessing step.
+
+        Args:
+            session (Session): Session object.
+            raw_sumstats_path (str): Input raw summary stats path.
+            out_sumstats_path (str): Output summary stats path.
+        """
         # Process summary stats.
         (
-            FinnGenSummaryStats.from_source(
-                self.session.spark, raw_file=self.raw_sumstats_path
-            )
-            .df.write.mode(self.session.write_mode)
-            .parquet(self.out_sumstats_path)
+            FinnGenSummaryStats.from_source(session.spark, raw_file=raw_sumstats_path)
+            .df.write.mode(session.write_mode)
+            .parquet(out_sumstats_path)
         )
diff --git a/src/otg/gene_index.py b/src/otg/gene_index.py
index 4a2cc34c1..c4843572e 100644
--- a/src/otg/gene_index.py
+++ b/src/otg/gene_index.py
@@ -1,35 +1,31 @@
 """Step to generate gene index dataset."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.datasource.open_targets.target import OpenTargetsTarget
 
 
-@dataclass
 class GeneIndexStep:
     """Gene index step.
 
     This step generates a gene index dataset from an Open Targets Platform target dataset.
-
-    Attributes:
-        session (Session): Session object.
-        target_path (str): Open targets Platform target dataset path.
-        gene_index_path (str): Output gene index path.
     """
 
-    session: Session = MISSING
-    target_path: str = MISSING
-    gene_index_path: str = MISSING
-
-    def __post_init__(self: GeneIndexStep) -> None:
-        """Run step."""
-        # Extract
-        platform_target = self.session.spark.read.parquet(self.target_path)
+    def __init__(
+        self,
+        session: Session,
+        target_path: str,
+        gene_index_path: str,
+    ) -> None:
+        """Initialize step.
+
+        Args:
+            session (Session): Session object.
+            target_path (str): Input Open Targets Platform target dataset path.
+            gene_index_path (str): Output gene index dataset path.
+        """
+        platform_target = session.spark.read.parquet(target_path)
         # Transform
         gene_index = OpenTargetsTarget.as_gene_index(platform_target)
         # Load
-        gene_index.df.write.mode(self.session.write_mode).parquet(self.gene_index_path)
+        gene_index.df.write.mode(session.write_mode).parquet(gene_index_path)
diff --git a/src/otg/gwas_catalog_ingestion.py b/src/otg/gwas_catalog_ingestion.py
index d47f82337..07c599ddc 100644
--- a/src/otg/gwas_catalog_ingestion.py
+++ b/src/otg/gwas_catalog_ingestion.py
@@ -1,10 +1,6 @@
 """Step to process GWAS Catalog associations and study table."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.dataset.variant_annotation import VariantAnnotation
 from otg.datasource.gwas_catalog.associations import (
@@ -17,54 +13,55 @@
 from otg.datasource.gwas_catalog.study_splitter import GWASCatalogStudySplitter
 
 
-@dataclass
 class GWASCatalogIngestionStep:
     """GWAS Catalog ingestion step to extract GWASCatalog Study and StudyLocus tables.
 
-    !!!note This step currently only processes the GWAS Catalog curated list of top hits.
-
-    Attributes:
-        session (Session): Session object.
-        catalog_study_files (list[str]): List of raw GWAS catalog studies file.
-        catalog_ancestry_files (list[str]): List of raw ancestry annotations files from GWAS Catalog.
-        catalog_sumstats_lut (str): GWAS Catalog summary statistics lookup table.
-        catalog_associations_file (str): Raw GWAS catalog associations file.
-        gwas_catalog_study_curation_file (str | None): file of the curation table. Optional.
-        variant_annotation_path (str): Input variant annotation path.
-        catalog_studies_out (str): Output GWAS catalog studies path.
-        catalog_associations_out (str): Output GWAS catalog associations path.
-        inclusion_list_path (str | None): optional inclusion list (parquet)
+    !!! note This step currently only processes the GWAS Catalog curated list of top hits.
     """
 
-    session: Session = MISSING
-    catalog_study_files: list[str] = MISSING
-    catalog_ancestry_files: list[str] = MISSING
-    catalog_sumstats_lut: str = MISSING
-    catalog_associations_file: str = MISSING
-    gwas_catalog_study_curation_file: str | None = None
-    variant_annotation_path: str = MISSING
-    catalog_studies_out: str = MISSING
-    catalog_associations_out: str = MISSING
-    inclusion_list_path: str | None = None
+    def __init__(
+        self,
+        session: Session,
+        catalog_study_files: list[str],
+        catalog_ancestry_files: list[str],
+        catalog_sumstats_lut: str,
+        catalog_associations_file: str,
+        variant_annotation_path: str,
+        catalog_studies_out: str,
+        catalog_associations_out: str,
+        gwas_catalog_study_curation_file: str | None = None,
+        inclusion_list_path: str | None = None,
+    ) -> None:
+        """Run step.
 
-    def __post_init__(self: GWASCatalogIngestionStep) -> None:
-        """Run step."""
+        Args:
+            session (Session): Session object.
+            catalog_study_files (list[str]): List of raw GWAS catalog studies file.
+            catalog_ancestry_files (list[str]): List of raw ancestry annotations files from GWAS Catalog.
+            catalog_sumstats_lut (str): GWAS Catalog summary statistics lookup table.
+            catalog_associations_file (str): Raw GWAS catalog associations file.
+            variant_annotation_path (str): Input variant annotation path.
+            catalog_studies_out (str): Output GWAS catalog studies path.
+            catalog_associations_out (str): Output GWAS catalog associations path.
+            gwas_catalog_study_curation_file (str | None): file of the curation table. Optional.
+            inclusion_list_path (str | None): optional inclusion list (parquet)
+        """
         # Extract
-        va = VariantAnnotation.from_parquet(self.session, self.variant_annotation_path)
-        catalog_studies = self.session.spark.read.csv(
-            list(self.catalog_study_files), sep="\t", header=True
+        va = VariantAnnotation.from_parquet(session, variant_annotation_path)
+        catalog_studies = session.spark.read.csv(
+            list(catalog_study_files), sep="\t", header=True
         )
-        ancestry_lut = self.session.spark.read.csv(
-            list(self.catalog_ancestry_files), sep="\t", header=True
+        ancestry_lut = session.spark.read.csv(
+            list(catalog_ancestry_files), sep="\t", header=True
         )
-        sumstats_lut = self.session.spark.read.csv(
-            self.catalog_sumstats_lut, sep="\t", header=False
+        sumstats_lut = session.spark.read.csv(
+            catalog_sumstats_lut, sep="\t", header=False
         )
-        catalog_associations = self.session.spark.read.csv(
-            self.catalog_associations_file, sep="\t", header=True
+        catalog_associations = session.spark.read.csv(
+            catalog_associations_file, sep="\t", header=True
         ).persist()
         gwas_catalog_study_curation = read_curation_table(
-            self.gwas_catalog_study_curation_file, self.session
+            gwas_catalog_study_curation_file, session
         )
 
         # Transform
@@ -76,18 +73,14 @@ def __post_init__(self: GWASCatalogIngestionStep) -> None:
         )
 
         # if inclusion list is provided apply filter:
-        if self.inclusion_list_path:
-            inclusion_list = self.session.spark.read.parquet(
-                self.inclusion_list_path, sep="\t", header=True
+        if inclusion_list_path:
+            inclusion_list = session.spark.read.parquet(
+                inclusion_list_path, sep="\t", header=True
             )
 
             study_index = study_index.apply_inclusion_list(inclusion_list)
             study_locus = study_locus.apply_inclusion_list(inclusion_list)
 
         # Load
-        study_index.df.write.mode(self.session.write_mode).parquet(
-            self.catalog_studies_out
-        )
-        study_locus.df.write.mode(self.session.write_mode).parquet(
-            self.catalog_associations_out
-        )
+        study_index.df.write.mode(session.write_mode).parquet(catalog_studies_out)
+        study_locus.df.write.mode(session.write_mode).parquet(catalog_associations_out)
diff --git a/src/otg/gwas_catalog_study_curation.py b/src/otg/gwas_catalog_study_curation.py
index 97f3a0b0a..aa4b713b0 100644
--- a/src/otg/gwas_catalog_study_curation.py
+++ b/src/otg/gwas_catalog_study_curation.py
@@ -1,10 +1,6 @@
 """Step to update GWAS Catalog study curation file based on newly released GWAS Catalog dataset."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.datasource.gwas_catalog.study_index import (
     StudyIndexGWASCatalogParser,
@@ -12,39 +8,39 @@
 )
 
 
-@dataclass
 class GWASCatalogStudyCurationStep:
-    """Create an updated curation table for GWAS Catalog study table.
-
-    Attributes:
-        session (Session): Session object.
-        catalog_study_files (list[str]): List of raw GWAS catalog studies file.
-        catalog_ancestry_files (list[str]): List of raw ancestry annotations files from GWAS Catalog.
-        catalog_sumstats_lut (str): GWAS Catalog summary statistics lookup table.
-        gwas_catalog_study_curation_file (str | None): Path to the original curation table. Optinal
-        gwas_catalog_study_curation_out (str): Path for the updated curation table.
-    """
-
-    session: Session = MISSING
-    catalog_study_files: list[str] = MISSING
-    catalog_ancestry_files: list[str] = MISSING
-    catalog_sumstats_lut: str = MISSING
-    gwas_catalog_study_curation_file: str | None = MISSING
-    gwas_catalog_study_curation_out: str = MISSING
-
-    def __post_init__(self: GWASCatalogStudyCurationStep) -> None:
-        """Run step."""
-        catalog_studies = self.session.spark.read.csv(
-            list(self.catalog_study_files), sep="\t", header=True
+    """Annotate GWAS Catalog studies with additional curation and create a curation backlog."""
+
+    def __init__(
+        self,
+        session: Session,
+        catalog_study_files: list[str],
+        catalog_ancestry_files: list[str],
+        catalog_sumstats_lut: str,
+        gwas_catalog_study_curation_out: str,
+        gwas_catalog_study_curation_file: str | None,
+    ) -> None:
+        """Run step to annotate and create backlog.
+
+        Args:
+            session (Session): Session object.
+            catalog_study_files (list[str]): List of raw GWAS catalog studies file.
+            catalog_ancestry_files (list[str]): List of raw ancestry annotations files from GWAS Catalog.
+            catalog_sumstats_lut (str): GWAS Catalog summary statistics lookup table.
+            gwas_catalog_study_curation_out (str): Path for the updated curation table.
+            gwas_catalog_study_curation_file (str | None): Path to the original curation table. Optinal
+        """
+        catalog_studies = session.spark.read.csv(
+            list(catalog_study_files), sep="\t", header=True
         )
-        ancestry_lut = self.session.spark.read.csv(
-            list(self.catalog_ancestry_files), sep="\t", header=True
+        ancestry_lut = session.spark.read.csv(
+            list(catalog_ancestry_files), sep="\t", header=True
         )
-        sumstats_lut = self.session.spark.read.csv(
-            self.catalog_sumstats_lut, sep="\t", header=False
+        sumstats_lut = session.spark.read.csv(
+            catalog_sumstats_lut, sep="\t", header=False
         )
         gwas_catalog_study_curation = read_curation_table(
-            self.gwas_catalog_study_curation_file, self.session
+            gwas_catalog_study_curation_file, session
         )
 
         # Process GWAS Catalog studies and get list of studies for curation:
@@ -58,5 +54,5 @@ def __post_init__(self: GWASCatalogStudyCurationStep) -> None:
             .extract_studies_for_curation(gwas_catalog_study_curation)
             # Save table:
             .toPandas()
-            .to_csv(self.gwas_catalog_study_curation_out, sep="\t", index=False)
+            .to_csv(gwas_catalog_study_curation_out, sep="\t", index=False)
         )
diff --git a/src/otg/gwas_catalog_study_inclusion.py b/src/otg/gwas_catalog_study_inclusion.py
index 8dd95d85f..06cf1a62d 100644
--- a/src/otg/gwas_catalog_study_inclusion.py
+++ b/src/otg/gwas_catalog_study_inclusion.py
@@ -1,10 +1,8 @@
 """Step to generate an GWAS Catalog study identifier inclusion and exclusion list."""
 from __future__ import annotations
 
-from dataclasses import dataclass
 from typing import TYPE_CHECKING
 
-from omegaconf import MISSING
 from pyspark.sql import functions as f
 
 from otg.common.session import Session
@@ -23,38 +21,8 @@
     from pyspark.sql import Column, DataFrame
 
 
-@dataclass
-class GWASCatalogInclusionGenerator:
-    """Generate GWAS Catalog studies eligible for ingestion based on curation and the provided criteria.
-
-    Attributes:
-        session (Session): Session object.
-        catalog_study_files (list[str]): List of raw GWAS catalog studies file.
-        catalog_ancestry_files (list[str]): List of raw ancestry annotations files from GWAS Catalog.
-        catalog_associations_file (str): Raw GWAS catalog associations file.
-        variant_annotation_path (str): Input variant annotation path.
-        gwas_catalog_study_curation_file (str | None): file of the curation table. Optional.
-        harmonised_study_file (str): file with the list of all successfully harmonized summary statistics.
-        criteria (str): inclusion criteria depending on the dataset ("summary_stats" or "curation")
-        inclusion_list_path (str): output file name of the list of study identifiers that can be ingested.
-        exclusion_list_path (str): output file name of the list of study identifiers that excluded from ingestion.
-    """
-
-    session: Session = MISSING
-    # GWAS Catalog sources:
-    catalog_study_files: list[str] = MISSING
-    catalog_ancestry_files: list[str] = MISSING
-    catalog_associations_file: str = MISSING
-    # Custom curation:
-    gwas_catalog_study_curation_file: str | None = None
-    # GnomAD source:
-    variant_annotation_path: str = MISSING
-    # Dynamic sources:
-    harmonised_study_file: str = MISSING
-    criteria: str = MISSING
-    # Output:
-    inclusion_list_path: str = MISSING
-    exclusion_list_path: str = MISSING
+class GWASCatalogStudyInclusionGenerator:
+    """GWAS Catalog study eligibility for ingestion based on curation and the provided criteria."""
 
     @staticmethod
     def flag_eligible_studies(
@@ -113,34 +81,50 @@ def process_harmonised_list(studies: list[str], session: Session) -> DataFrame:
         """
         return session.spark.createDataFrame([{"_c0": path} for path in studies])
 
+    @staticmethod
     def get_gwas_catalog_study_index(
-        self: GWASCatalogInclusionGenerator,
+        session: Session,
+        variant_annotation_path: str,
+        catalog_study_files: list[str],
+        catalog_ancestry_files: list[str],
+        harmonised_study_file: str,
+        catalog_associations_file: str,
+        gwas_catalog_study_curation_file: str,
     ) -> StudyIndexGWASCatalog:
         """Return GWAS Catalog study index.
 
+        Args:
+            session (Session): Session object.
+            variant_annotation_path (str): Input variant annotation path.
+            catalog_study_files (list[str]): List of raw GWAS catalog studies file.
+            catalog_ancestry_files (list[str]): List of raw ancestry annotations files from GWAS Catalog.
+            harmonised_study_file (str): GWAS Catalog summary statistics lookup table.
+            catalog_associations_file (str): Raw GWAS catalog associations file.
+            gwas_catalog_study_curation_file (str): file of the curation table. Optional.
+
         Returns:
             StudyIndexGWASCatalog: Completely processed and fully annotated study index.
         """
         # Extract
-        va = VariantAnnotation.from_parquet(self.session, self.variant_annotation_path)
-        catalog_studies = self.session.spark.read.csv(
-            list(self.catalog_study_files), sep="\t", header=True
+        va = VariantAnnotation.from_parquet(session, variant_annotation_path)
+        catalog_studies = session.spark.read.csv(
+            list(catalog_study_files), sep="\t", header=True
         )
-        ancestry_lut = self.session.spark.read.csv(
-            list(self.catalog_ancestry_files), sep="\t", header=True
+        ancestry_lut = session.spark.read.csv(
+            list(catalog_ancestry_files), sep="\t", header=True
         )
-        sumstats_lut = self.session.spark.read.csv(
-            self.harmonised_study_file, sep="\t", header=False
+        sumstats_lut = session.spark.read.csv(
+            harmonised_study_file, sep="\t", header=False
         )
-        catalog_associations = self.session.spark.read.csv(
-            self.catalog_associations_file, sep="\t", header=True
+        catalog_associations = session.spark.read.csv(
+            catalog_associations_file, sep="\t", header=True
         ).persist()
         gwas_catalog_study_curation = read_curation_table(
-            self.gwas_catalog_study_curation_file, self.session
+            gwas_catalog_study_curation_file, session
         )
 
         # Transform
-        study_index, study_locus = GWASCatalogStudySplitter.split(
+        study_index, _ = GWASCatalogStudySplitter.split(
             StudyIndexGWASCatalogParser.from_source(
                 catalog_studies,
                 ancestry_lut,
@@ -151,20 +135,53 @@ def get_gwas_catalog_study_index(
 
         return study_index
 
-    def __post_init__(self: GWASCatalogInclusionGenerator) -> None:
-        """Run step."""
+    def __init__(
+        self,
+        session: Session,
+        catalog_study_files: list[str],
+        catalog_ancestry_files: list[str],
+        catalog_associations_file: str,
+        gwas_catalog_study_curation_file: str,
+        variant_annotation_path: str,
+        harmonised_study_file: str,
+        criteria: str,
+        inclusion_list_path: str,
+        exclusion_list_path: str,
+    ) -> None:
+        """Run step.
+
+        Args:
+            session (Session): Session object.
+            catalog_study_files (list[str]): List of raw GWAS catalog studies file.
+            catalog_ancestry_files (list[str]): List of raw ancestry annotations files from GWAS Catalog.
+            catalog_associations_file (str): Raw GWAS catalog associations file.
+            gwas_catalog_study_curation_file (str): file of the curation table. Optional.
+            variant_annotation_path (str): Input variant annotation path.
+            harmonised_study_file (str): GWAS Catalog summary statistics lookup table.
+            criteria (str): name of the filter set to be applied.
+            inclusion_list_path (str): Output path for the inclusion list.
+            exclusion_list_path (str): Output path for the exclusion list.
+        """
         # Create study index:
-        study_index = self.get_gwas_catalog_study_index()
+        study_index = self.get_gwas_catalog_study_index(
+            session,
+            variant_annotation_path,
+            catalog_study_files,
+            catalog_ancestry_files,
+            harmonised_study_file,
+            catalog_associations_file,
+            gwas_catalog_study_curation_file,
+        )
 
         # Get study indices for inclusion:
-        flagged_studies = self.flag_eligible_studies(study_index, self.criteria)
+        flagged_studies = self.flag_eligible_studies(study_index, criteria)
 
         # Output inclusion list:
         eligible = (
             flagged_studies.filter(f.col("isEligible")).select("studyId").persist()
         )
-        eligible.write.mode(self.session.write_mode).parquet(self.inclusion_list_path)
+        eligible.write.mode(session.write_mode).parquet(inclusion_list_path)
 
         # Output exclusion list:
         excluded = flagged_studies.filter(~f.col("isEligible")).persist()
-        excluded.write.mode(self.session.write_mode).parquet(self.exclusion_list_path)
+        excluded.write.mode(session.write_mode).parquet(exclusion_list_path)
diff --git a/src/otg/gwas_catalog_sumstat_preprocess.py b/src/otg/gwas_catalog_sumstat_preprocess.py
index 682aed1cf..7ef6a1cd0 100644
--- a/src/otg/gwas_catalog_sumstat_preprocess.py
+++ b/src/otg/gwas_catalog_sumstat_preprocess.py
@@ -1,36 +1,25 @@
 """Step to generate variant annotation dataset."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.datasource.gwas_catalog.summary_statistics import GWASCatalogSummaryStatistics
 
 
-@dataclass
 class GWASCatalogSumstatsPreprocessStep:
-    """Step to preprocess GWAS Catalog harmonised summary stats.
-
-    Attributes:
-        session (Session): Session object.
-        raw_sumstats_path (str): Input raw GWAS Catalog summary statistics path.
-        out_sumstats_path (str): Output GWAS Catalog summary statistics path.
-    """
-
-    session: Session = MISSING
-    raw_sumstats_path: str = MISSING
-    out_sumstats_path: str = MISSING
-
-    def __post_init__(self: GWASCatalogSumstatsPreprocessStep) -> None:
-        """Run step."""
-        # Extract
-        self.session.logger.info(self.raw_sumstats_path)
-        self.session.logger.info(self.out_sumstats_path)
-
+    """Step to preprocess GWAS Catalog harmonised summary stats."""
+
+    def __init__(
+        self, session: Session, raw_sumstats_path: str, out_sumstats_path: str
+    ) -> None:
+        """Run step to preprocess GWAS Catalog harmonised summary stats and produce SummaryStatistics dataset.
+
+        Args:
+            session (Session): Session object.
+            raw_sumstats_path (str): Input GWAS Catalog harmonised summary stats path.
+            out_sumstats_path (str): Output SummaryStatistics dataset path.
+        """
         # Processing dataset:
         GWASCatalogSummaryStatistics.from_gwas_harmonized_summary_stats(
-            self.session.spark, self.raw_sumstats_path
-        ).df.write.mode(self.session.write_mode).parquet(self.out_sumstats_path)
-        self.session.logger.info("Processing dataset successfully completed.")
+            session.spark, raw_sumstats_path
+        ).df.write.mode(session.write_mode).parquet(out_sumstats_path)
+        session.logger.info("Processing dataset successfully completed.")
diff --git a/src/otg/l2g.py b/src/otg/l2g.py
index 2c61f58eb..10f9c1224 100644
--- a/src/otg/l2g.py
+++ b/src/otg/l2g.py
@@ -1,12 +1,10 @@
 """Step to run Locus to Gene either for inference or for training."""
 from __future__ import annotations
 
-from dataclasses import dataclass, field
 from typing import Any
 
 import pyspark.sql.functions as f
 import sklearn
-from omegaconf import MISSING
 from xgboost.spark import SparkXGBClassifier
 
 from otg.common.session import Session
@@ -21,111 +19,66 @@
 from otg.method.l2g.trainer import LocusToGeneTrainer
 
 
-@dataclass
 class LocusToGeneStep:
-    """Locus to gene step.
-
-    Attributes:
-        session (Session): Session object.
-        extended_spark_conf (dict[str, str] | None): Extended Spark configuration.
-        run_mode (str): One of "train" or "predict".
-        wandb_run_name (str | None): Name of the run to be tracked on W&B.
-        perform_cross_validation (bool): Whether to perform cross validation.
-        model_path (str | None): Path to save the model.
-        predictions_path (str | None): Path to save the predictions.
-        credible_set_path (str): Path to credible set Parquet files.
-        variant_gene_path (str): Path to variant to gene Parquet files.
-        colocalisation_path (str): Path to colocalisation Parquet files.
-        study_index_path (str): Path to study index Parquet files.
-        study_locus_overlap_path (str): Path to study locus overlap Parquet files.
-        gold_standard_curation_path (str | None): Path to gold standard curation JSON files.
-        gene_interactions_path (str | None): Path to gene interactions Parquet files.
-        features_list (list[str]): List of features to use.
-        hyperparameters (dict): Hyperparameters for the model.
-    """
-
-    extended_spark_conf: dict[str, str] | None = None
-
-    session: Session = MISSING
-    run_mode: str = MISSING
-    wandb_run_name: str | None = None
-    perform_cross_validation: bool = False
-    model_path: str = MISSING
-    predictions_path: str = MISSING
-    credible_set_path: str = MISSING
-    variant_gene_path: str = MISSING
-    colocalisation_path: str = MISSING
-    study_index_path: str = MISSING
-    gold_standard_curation_path: str = MISSING
-    gene_interactions_path: str = MISSING
-    features_list: list[str] = field(
-        default_factory=lambda: [
-            # average distance of all tagging variants to gene TSS
-            "distanceTssMean",
-            # # minimum distance of all tagging variants to gene TSS
-            "distanceTssMinimum",
-            # # maximum vep consequence score of the locus 95% credible set among all genes in the vicinity
-            "vepMaximumNeighborhood",
-            # # maximum vep consequence score of the locus 95% credible set split by gene
-            "vepMaximum",
-            # max clpp for each (study, locus, gene) aggregating over all eQTLs
-            "eqtlColocClppMaximum",
-            # max clpp for each (study, locus) aggregating over all eQTLs
-            "eqtlColocClppMaximumNeighborhood",
-            # max clpp for each (study, locus, gene) aggregating over all pQTLs
-            # "pqtlColocClppMaximum",
-            # max clpp for each (study, locus) aggregating over all pQTLs
-            # "pqtlColocClppMaximumNeighborhood",
-            # max clpp for each (study, locus, gene) aggregating over all sQTLs
-            # "sqtlColocClppMaximum",
-            # max clpp for each (study, locus) aggregating over all sQTLs
-            # "sqtlColocClppMaximumNeighborhood",
-            # # max log-likelihood ratio value for each (study, locus, gene) aggregating over all eQTLs
-            # "eqtlColocLlrLocalMaximum",
-            # # max log-likelihood ratio value for each (study, locus) aggregating over all eQTLs
-            # "eqtlColocLlpMaximumNeighborhood",
-            # # max log-likelihood ratio value for each (study, locus, gene) aggregating over all pQTLs
-            # "pqtlColocLlrLocalMaximum",
-            # # max log-likelihood ratio value for each (study, locus) aggregating over all pQTLs
-            # "pqtlColocLlpMaximumNeighborhood",
-            # # max log-likelihood ratio value for each (study, locus, gene) aggregating over all sQTLs
-            # "sqtlColocLlrLocalMaximum",
-            # # max log-likelihood ratio value for each (study, locus) aggregating over all sQTLs
-            # "sqtlColocLlpMaximumNeighborhood",
-        ]
-    )
-    hyperparameters: dict[str, Any] = field(
-        default_factory=lambda: {
-            "max_depth": 5,
-            "loss_function": "binary:logistic",
-        }
-    )
-
-    def __post_init__(self: LocusToGeneStep) -> None:
+    """Locus to gene step."""
+
+    def __init__(
+        self,
+        session: Session,
+        run_mode: str,
+        model_path: str,
+        predictions_path: str,
+        credible_set_path: str,
+        variant_gene_path: str,
+        colocalisation_path: str,
+        study_index_path: str,
+        gold_standard_curation_path: str,
+        gene_interactions_path: str,
+        features_list: list[str],
+        hyperparameters: dict[str, Any],
+        wandb_run_name: str | None = None,
+        perform_cross_validation: bool = False,
+    ) -> None:
         """Run step.
 
+        Args:
+            session (Session): Session object.
+            run_mode (str): One of "train" or "predict".
+            model_path (str): Path to save the model.
+            predictions_path (str): Path to save the predictions.
+            credible_set_path (str): Path to credible set Parquet files.
+            variant_gene_path (str): Path to variant to gene Parquet files.
+            colocalisation_path (str): Path to colocalisation Parquet files.
+            study_index_path (str): Path to study index Parquet files.
+            gold_standard_curation_path (str): Path to gold standard curation JSON files.
+            gene_interactions_path (str): Path to gene interactions Parquet files.
+            features_list (list[str]): List of features to use.
+            hyperparameters (dict[str, Any]): Hyperparameters for the model.
+            wandb_run_name (str | None): Name of the run to be tracked on W&B.
+            perform_cross_validation (bool): Whether to perform cross validation.
+
         Raises:
             ValueError: if run_mode is not one of "train" or "predict".
         """
         print("Sci-kit learn version: ", sklearn.__version__)  # noqa: T201
-        if self.run_mode not in ["train", "predict"]:
+        if run_mode not in ["train", "predict"]:
             raise ValueError(
-                f"run_mode must be one of 'train' or 'predict', got {self.run_mode}"
+                f"run_mode must be one of 'train' or 'predict', got {run_mode}"
             )
         # Load common inputs
         credible_set = StudyLocus.from_parquet(
-            self.session, self.credible_set_path, recursiveFileLookup=True
+            session, credible_set_path, recursiveFileLookup=True
         )
         studies = StudyIndex.from_parquet(
-            self.session, self.study_index_path, recursiveFileLookup=True
+            session, study_index_path, recursiveFileLookup=True
         )
-        v2g = V2G.from_parquet(self.session, self.variant_gene_path)
-        coloc = Colocalisation.from_parquet(self.session, self.colocalisation_path)
+        v2g = V2G.from_parquet(session, variant_gene_path)
+        coloc = Colocalisation.from_parquet(session, colocalisation_path)
 
-        if self.run_mode == "train":
+        if run_mode == "train":
             # Process gold standard and L2G features
-            gs_curation = self.session.spark.read.json(self.gold_standard_curation_path)
-            interactions = self.session.spark.read.parquet(self.gene_interactions_path)
+            gs_curation = session.spark.read.json(gold_standard_curation_path)
+            interactions = session.spark.read.parquet(gene_interactions_path)
             study_locus_overlap = StudyLocus(
                 # We just extract overlaps of associations in the gold standard. This parsing is a duplication of the one in the gold standard curation,
                 # but we need to do it here to be able to parse gold standards later
@@ -158,7 +111,7 @@ def __post_init__(self: LocusToGeneStep) -> None:
             )
 
             fm = L2GFeatureMatrix.generate_features(
-                features_list=self.features_list,
+                features_list=features_list,
                 study_locus=credible_set,
                 study_index=studies,
                 variant_gene=v2g,
@@ -185,11 +138,11 @@ def __post_init__(self: LocusToGeneStep) -> None:
                 max_depth=5,
             )
             l2g_model = LocusToGeneModel(
-                features_list=list(self.features_list), estimator=estimator
+                features_list=list(features_list), estimator=estimator
             )
-            if self.perform_cross_validation:
+            if perform_cross_validation:
                 # Perform cross validation to extract what are the best hyperparameters
-                cv_folds = self.hyperparameters.get("cross_validation_folds", 5)
+                cv_folds = hyperparameters.get("cross_validation_folds", 5)
                 LocusToGeneTrainer.cross_validate(
                     l2g_model=l2g_model,
                     data=data,
@@ -200,23 +153,21 @@ def __post_init__(self: LocusToGeneStep) -> None:
                 LocusToGeneTrainer.train(
                     data=data,
                     l2g_model=l2g_model,
-                    features_list=list(self.features_list),
-                    model_path=self.model_path,
+                    features_list=list(features_list),
+                    model_path=model_path,
                     evaluate=True,
-                    wandb_run_name=self.wandb_run_name,
-                    **self.hyperparameters,
+                    wandb_run_name=wandb_run_name,
+                    **hyperparameters,
                 )
-                self.session.logger.info(self.model_path)
+                session.logger.info(model_path)
 
-        if self.run_mode == "predict":
-            if not self.model_path or not self.predictions_path:
+        if run_mode == "predict":
+            if not model_path or not predictions_path:
                 raise ValueError(
                     "model_path and predictions_path must be set for predict mode."
                 )
             predictions = L2GPrediction.from_credible_set(
-                self.model_path, self.features_list, credible_set, studies, v2g, coloc
-            )
-            predictions.df.write.mode(self.session.write_mode).parquet(
-                self.predictions_path
+                model_path, features_list, credible_set, studies, v2g, coloc
             )
-            self.session.logger.info(self.predictions_path)
+            predictions.df.write.mode(session.write_mode).parquet(predictions_path)
+            session.logger.info(predictions_path)
diff --git a/src/otg/ld_based_clumping.py b/src/otg/ld_based_clumping.py
index 6158f9d36..127dfdaa6 100644
--- a/src/otg/ld_based_clumping.py
+++ b/src/otg/ld_based_clumping.py
@@ -1,17 +1,12 @@
 """Step to apply linkageg based clumping on study-locus dataset."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.dataset.ld_index import LDIndex
 from otg.dataset.study_index import StudyIndex
 from otg.dataset.study_locus import StudyLocus
 
 
-@dataclass
 class LdBasedClumpingStep:
     """Step to perform LD-based clumping on study locus dataset.
 
@@ -19,26 +14,28 @@ class LdBasedClumpingStep:
     That's why the study index and the ld index is required for this step. Study loci are flaggged
     in the resulting dataset, which can be explained by a more significant association
     from the same study.
-
-    Attributes:
-        session (Session): Session object
-        study_locus_input_path (str): Path to the input study locus.
-        study_index_path (str): Path to the study index.
-        ld_index_path (str): Path to the LD index.
-        clumped_study_locus_output_path (str): path of the resulting, clumped study-locus dataset.
     """
 
-    session: Session = MISSING
-    study_locus_input_path: str = MISSING
-    study_index_path: str = MISSING
-    ld_index_path: str = MISSING
-    clumped_study_locus_output_path: str = MISSING
-
-    def __post_init__(self: LdBasedClumpingStep) -> None:
-        """Run clumping step."""
-        study_locus = StudyLocus.from_parquet(self.session, self.study_locus_input_path)
-        ld_index = LDIndex.from_parquet(self.session, self.ld_index_path)
-        study_index = StudyIndex.from_parquet(self.session, self.study_index_path)
+    def __init__(
+        self,
+        session: Session,
+        study_locus_input_path: str,
+        study_index_path: str,
+        ld_index_path: str,
+        clumped_study_locus_output_path: str,
+    ) -> None:
+        """Run LD-based clumping step.
+
+        Args:
+            session (Session): Session object.
+            study_locus_input_path (str): Path to the input study locus.
+            study_index_path (str): Path to the study index.
+            ld_index_path (str): Path to the LD index.
+            clumped_study_locus_output_path (str): path of the resulting, clumped study-locus dataset.
+        """
+        study_locus = StudyLocus.from_parquet(session, study_locus_input_path)
+        ld_index = LDIndex.from_parquet(session, ld_index_path)
+        study_index = StudyIndex.from_parquet(session, study_index_path)
 
         (
             study_locus
@@ -46,6 +43,6 @@ def __post_init__(self: LdBasedClumpingStep) -> None:
             .annotate_ld(study_index, ld_index)
             .clump()
             # Save result:
-            .df.write.mode(self.session.write_mode)
-            .parquet(self.clumped_study_locus_output_path)
+            .df.write.mode(session.write_mode)
+            .parquet(clumped_study_locus_output_path)
         )
diff --git a/src/otg/ld_index.py b/src/otg/ld_index.py
index 54ab93066..d88839852 100644
--- a/src/otg/ld_index.py
+++ b/src/otg/ld_index.py
@@ -1,41 +1,33 @@
 """Step to dump a filtered version of a LD matrix (block matrix) as Parquet files."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
 import hail as hl
-from omegaconf import MISSING
 
 from otg.common.session import Session
 from otg.datasource.gnomad.ld import GnomADLDMatrix
 
 
-@dataclass
 class LDIndexStep:
     """LD index step.
 
     !!! warning "This step is resource intensive"
         Suggested params: high memory machine, 5TB of boot disk, no SSDs.
-
-    Attributes:
-        session (Session): Session object.
-        min_r2 (float): Minimum r2 to consider when considering variants within a window.
-        ld_index_out (str): Output LD index path.
     """
 
-    session: Session = MISSING
-
-    min_r2: float = 0.5
-    ld_index_out: str = MISSING
+    def __init__(self, session: Session, min_r2: float, ld_index_out: str) -> None:
+        """Run step.
 
-    def __post_init__(self: LDIndexStep) -> None:
-        """Run step."""
-        hl.init(sc=self.session.spark.sparkContext, log="/dev/null")
+        Args:
+            session (Session): Session object.
+            min_r2 (float): Minimum r2 to consider when considering variants within a window.
+            ld_index_out (str): Output LD index path.
+        """
+        hl.init(sc=session.spark.sparkContext, log="/dev/null")
         (
             GnomADLDMatrix()
-            .as_ld_index(self.min_r2)
+            .as_ld_index(min_r2)
             .df.write.partitionBy("chromosome")
-            .mode(self.session.write_mode)
-            .parquet(self.ld_index_out)
+            .mode(session.write_mode)
+            .parquet(ld_index_out)
         )
-        self.session.logger.info(self.ld_index_out)
+        session.logger.info(ld_index_out)
diff --git a/src/otg/overlaps.py b/src/otg/overlaps.py
index 74a8c6d58..2949f1fd8 100644
--- a/src/otg/overlaps.py
+++ b/src/otg/overlaps.py
@@ -1,17 +1,12 @@
 """Step to generate the dataset of overlapping studyLocus associations."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.dataset.study_index import StudyIndex
 from otg.dataset.study_locus import StudyLocus
 from otg.dataset.study_locus_overlap import StudyLocusOverlap
 
 
-@dataclass
 class OverlapsIndexStep:
     """StudyLocus overlaps step.
 
@@ -19,31 +14,31 @@ class OverlapsIndexStep:
     This dataset is defined to contain the overlapping signals between studyLocus associations once they have been clumped and fine-mapped.
 
     This step generates a dataset of overlapping studyLocus associations.
-
-    Attributes:
-        session (Session): Session object.
-        study_locus_path (str): Input study-locus path.
-        study_index_path (str): Input study index path to extract the type of study.
-        overlaps_index_out (str): Output overlaps index path.
     """
 
-    session: Session = MISSING
-    study_locus_path: str = MISSING
-    study_index_path: str = MISSING
-    overlaps_index_out: str = MISSING
-
-    def __post_init__(self: OverlapsIndexStep) -> None:
-        """Run step."""
+    def __init__(
+        self,
+        session: Session,
+        study_locus_path: str,
+        study_index_path: str,
+        overlaps_index_out: str,
+    ) -> None:
+        """Run step to calculate overlaps between studyLocus associations.
+
+        Args:
+            session (Session): Session object.
+            study_locus_path (str): Input studyLocus path.
+            study_index_path (str): Input studyIndex path.
+            overlaps_index_out (str): Output overlapsIndex path.
+        """
         # Extract
         study_locus = StudyLocus.from_parquet(
-            self.session, self.study_locus_path, recursiveFileLookup=True
+            session, study_locus_path, recursiveFileLookup=True
         )
         study_index = StudyIndex.from_parquet(
-            self.session, self.study_index_path, recursiveFileLookup=True
+            session, study_index_path, recursiveFileLookup=True
         )
         # Transform
         overlaps_index = StudyLocusOverlap.from_associations(study_locus, study_index)
         # Load
-        overlaps_index.df.write.mode(self.session.write_mode).parquet(
-            self.overlaps_index_out
-        )
+        overlaps_index.df.write.mode(session.write_mode).parquet(overlaps_index_out)
diff --git a/src/otg/pics.py b/src/otg/pics.py
index c326706cb..7634ab0fb 100644
--- a/src/otg/pics.py
+++ b/src/otg/pics.py
@@ -2,39 +2,32 @@
 
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.dataset.study_locus import StudyLocus
 from otg.method.pics import PICS
 
 
-@dataclass
 class PICSStep:
-    """PICS finemapping of LD-annotated StudyLocus.
-
-    Attributes:
-        session (Session): Session object.
-
-        study_locus_ld_annotated_in (str): Path to Study Locus with the LD information annotated
-        picsed_study_locus_out (str): Path to Study Locus after running PICS
-    """
-
-    session: Session = MISSING
-    study_locus_ld_annotated_in: str = MISSING
-    picsed_study_locus_out: str = MISSING
-
-    def __post_init__(self: PICSStep) -> None:
-        """Run step."""
+    """PICS finemapping of LD-annotated StudyLocus."""
+
+    def __init__(
+        self,
+        session: Session,
+        study_locus_ld_annotated_in: str,
+        picsed_study_locus_out: str,
+    ) -> None:
+        """Run PICS on LD annotated study-locus.
+
+        Args:
+            session (Session): Session object.
+            study_locus_ld_annotated_in (str): Input LD annotated study-locus path.
+            picsed_study_locus_out (str): Output PICSed study-locus path.
+        """
         # Extract
         study_locus_ld_annotated = StudyLocus.from_parquet(
-            self.session, self.study_locus_ld_annotated_in
+            session, study_locus_ld_annotated_in
         )
         # PICS
         picsed_sl = PICS.finemap(study_locus_ld_annotated).annotate_credible_sets()
         # Write
-        picsed_sl.df.write.mode(self.session.write_mode).parquet(
-            self.picsed_study_locus_out
-        )
+        picsed_sl.df.write.mode(session.write_mode).parquet(picsed_study_locus_out)
diff --git a/src/otg/ukbiobank.py b/src/otg/ukbiobank.py
index 3e1a25559..aadf8def9 100644
--- a/src/otg/ukbiobank.py
+++ b/src/otg/ukbiobank.py
@@ -2,39 +2,32 @@
 
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.datasource.ukbiobank.study_index import UKBiobankStudyIndex
 
 
-@dataclass
 class UKBiobankStep:
-    """UKBiobank study table ingestion step.
-
-    Attributes:
-        session (Session): Session object.
-        ukbiobank_manifest (str): UKBiobank manifest of studies.
-        ukbiobank_study_index_out (str): Output path for the UKBiobank study index dataset.
-    """
-
-    session: Session = MISSING
-    ukbiobank_manifest: str = MISSING
-    ukbiobank_study_index_out: str = MISSING
-
-    def __post_init__(self: UKBiobankStep) -> None:
-        """Run step."""
+    """UKBiobank study table ingestion step."""
+
+    def __init__(
+        self, session: Session, ukbiobank_manifest: str, ukbiobank_study_index_out: str
+    ) -> None:
+        """Run UKBB ingestion step.
+
+        Args:
+            session (Session): Session object.
+            ukbiobank_manifest (str): Input UKBiobank manifest path.
+            ukbiobank_study_index_out (str): Output UKBiobank study index path.
+        """
         # Read in the UKBiobank manifest tsv file.
-        df = self.session.spark.read.csv(
-            self.ukbiobank_manifest, sep="\t", header=True, inferSchema=True
+        df = session.spark.read.csv(
+            ukbiobank_manifest, sep="\t", header=True, inferSchema=True
         )
 
         # Parse the study index data.
         ukbiobank_study_index = UKBiobankStudyIndex.from_source(df)
 
         # Write the output.
-        ukbiobank_study_index.df.write.mode(self.session.write_mode).parquet(
-            self.ukbiobank_study_index_out
+        ukbiobank_study_index.df.write.mode(session.write_mode).parquet(
+            ukbiobank_study_index_out
         )
diff --git a/src/otg/v2g.py b/src/otg/v2g.py
index 61cf5fa79..b2f900cff 100644
--- a/src/otg/v2g.py
+++ b/src/otg/v2g.py
@@ -1,12 +1,9 @@
 """Step to generate variant annotation dataset."""
 from __future__ import annotations
 
-from dataclasses import dataclass, field
 from functools import reduce
-from typing import Dict, List
 
 import pyspark.sql.functions as f
-from omegaconf import MISSING
 
 from otg.common.Liftover import LiftOverSpark
 from otg.common.session import Session
@@ -17,7 +14,6 @@
 from otg.dataset.variant_index import VariantIndex
 
 
-@dataclass
 class V2GStep:
     """Variant-to-gene (V2G) step.
 
@@ -41,42 +37,41 @@ class V2GStep:
         v2g_path (str): Output V2G path.
     """
 
-    session: Session = MISSING
-    variant_index_path: str = MISSING
-    variant_annotation_path: str = MISSING
-    gene_index_path: str = MISSING
-    vep_consequences_path: str = MISSING
-    liftover_chain_file_path: str = MISSING
-    liftover_max_length_difference: int = 100
-    max_distance: int = 500_000
-    approved_biotypes: List[str] = field(
-        default_factory=lambda: [
-            "protein_coding",
-            "3prime_overlapping_ncRNA",
-            "antisense",
-            "bidirectional_promoter_lncRNA",
-            "IG_C_gene",
-            "IG_D_gene",
-            "IG_J_gene",
-            "IG_V_gene",
-            "lincRNA",
-            "macro_lncRNA",
-            "non_coding",
-            "sense_intronic",
-            "sense_overlapping",
-        ]
-    )
-    intervals: Dict[str, str] = field(default_factory=dict)
-    v2g_path: str = MISSING
+    def __init__(
+        self,
+        session: Session,
+        variant_index_path: str,
+        variant_annotation_path: str,
+        gene_index_path: str,
+        vep_consequences_path: str,
+        liftover_chain_file_path: str,
+        approved_biotypes: list[str],
+        interval_sources: dict[str, str],
+        v2g_path: str,
+        max_distance: int = 500_000,
+        liftover_max_length_difference: int = 100,
+    ) -> None:
+        """Run Variant-to-gene (V2G) step.
 
-    def __post_init__(self: V2GStep) -> None:
-        """Run step."""
+        Args:
+            session (Session): Session object.
+            variant_index_path (str): Input variant index path.
+            variant_annotation_path (str): Input variant annotation path.
+            gene_index_path (str): Input gene index path.
+            vep_consequences_path (str): Input VEP consequences path.
+            liftover_chain_file_path (str): Path to GRCh37 to GRCh38 chain file.
+            approved_biotypes (list[str]): List of approved biotypes.
+            interval_sources (dict[str, str]): Dictionary of interval sources.
+            v2g_path (str): Output V2G path.
+            max_distance (int): Maximum distance to consider.
+            liftover_max_length_difference (int): Maximum length difference for liftover.
+        """
         # Read
-        gene_index = GeneIndex.from_parquet(self.session, self.gene_index_path)
-        vi = VariantIndex.from_parquet(self.session, self.variant_index_path).persist()
-        va = VariantAnnotation.from_parquet(self.session, self.variant_annotation_path)
-        vep_consequences = self.session.spark.read.csv(
-            self.vep_consequences_path, sep="\t", header=True
+        gene_index = GeneIndex.from_parquet(session, gene_index_path)
+        vi = VariantIndex.from_parquet(session, variant_index_path).persist()
+        va = VariantAnnotation.from_parquet(session, variant_annotation_path)
+        vep_consequences = session.spark.read.csv(
+            vep_consequences_path, sep="\t", header=True
         ).select(
             f.element_at(f.split("Accession", r"/"), -1).alias(
                 "variantFunctionalConsequenceId"
@@ -88,12 +83,12 @@ def __post_init__(self: V2GStep) -> None:
         # Transform
         lift = LiftOverSpark(
             # lift over variants to hg38
-            self.liftover_chain_file_path,
-            self.liftover_max_length_difference,
+            liftover_chain_file_path,
+            liftover_max_length_difference,
         )
         gene_index_filtered = gene_index.filter_by_biotypes(
             # Filter gene index by approved biotypes to define V2G gene universe
-            list(self.approved_biotypes)
+            list(approved_biotypes)
         )
         va_slimmed = va.filter_by_variant_df(
             # Variant annotation reduced to the variant index to define V2G variant universe
@@ -105,15 +100,15 @@ def __post_init__(self: V2GStep) -> None:
                 # create interval instances by parsing each source
                 [
                     Intervals.from_source(
-                        self.session.spark, source_name, source_path, gene_index, lift
+                        session.spark, source_name, source_path, gene_index, lift
                     ).df
-                    for source_name, source_path in self.intervals.items()
+                    for source_name, source_path in interval_sources.items()
                 ],
             ),
             _schema=Intervals.get_schema(),
         )
         v2g_datasets = [
-            va_slimmed.get_distance_to_tss(gene_index_filtered, self.max_distance),
+            va_slimmed.get_distance_to_tss(gene_index_filtered, max_distance),
             va_slimmed.get_most_severe_vep_v2g(vep_consequences, gene_index_filtered),
             va_slimmed.get_plof_v2g(gene_index_filtered),
             intervals.v2g(vi),
@@ -129,6 +124,6 @@ def __post_init__(self: V2GStep) -> None:
         # Load
         (
             v2g.df.write.partitionBy("chromosome")
-            .mode(self.session.write_mode)
-            .parquet(self.v2g_path)
+            .mode(session.write_mode)
+            .parquet(v2g_path)
         )
diff --git a/src/otg/variant_annotation.py b/src/otg/variant_annotation.py
index 0573f8d50..42cc30f6d 100644
--- a/src/otg/variant_annotation.py
+++ b/src/otg/variant_annotation.py
@@ -1,38 +1,32 @@
 """Step to generate variant annotation dataset."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
 import hail as hl
-from omegaconf import MISSING
 
 from otg.common.session import Session
 from otg.datasource.gnomad.variants import GnomADVariants
 
 
-@dataclass
 class VariantAnnotationStep:
     """Variant annotation step.
 
     Variant annotation step produces a dataset of the type `VariantAnnotation` derived from gnomADs `gnomad.genomes.vX.X.X.sites.ht` Hail's table. This dataset is used to validate variants and as a source of annotation.
-
-    Attributes:
-        session (Session): Session object.
-        variant_annotation_path (str): Output variant annotation path.
     """
 
-    session: Session = MISSING
-    variant_annotation_path: str = MISSING
+    def __init__(self, session: Session, variant_annotation_path: str) -> None:
+        """Run Variant Annotation step.
 
-    def __post_init__(self: VariantAnnotationStep) -> None:
-        """Run step."""
+        Args:
+            session (Session): Session object.
+            variant_annotation_path (str): Variant annotation dataset path.
+        """
         # Initialise hail session.
-        hl.init(sc=self.session.spark.sparkContext, log="/dev/null")
+        hl.init(sc=session.spark.sparkContext, log="/dev/null")
         # Run variant annotation.
         variant_annotation = GnomADVariants().as_variant_annotation()
         # Write data partitioned by chromosome and position.
         (
-            variant_annotation.df.write.mode(self.session.write_mode).parquet(
-                self.variant_annotation_path
+            variant_annotation.df.write.mode(session.write_mode).parquet(
+                variant_annotation_path
             )
         )
diff --git a/src/otg/variant_index.py b/src/otg/variant_index.py
index 9b7b73ded..d08178a86 100644
--- a/src/otg/variant_index.py
+++ b/src/otg/variant_index.py
@@ -1,40 +1,37 @@
 """Step to generate variant index dataset."""
 from __future__ import annotations
 
-from dataclasses import dataclass
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.dataset.study_locus import StudyLocus
 from otg.dataset.variant_annotation import VariantAnnotation
 from otg.dataset.variant_index import VariantIndex
 
 
-@dataclass
 class VariantIndexStep:
     """Run variant index step to only variants in study-locus sets.
 
     Using a `VariantAnnotation` dataset as a reference, this step creates and writes a dataset of the type `VariantIndex` that includes only variants that have disease-association data with a reduced set of annotations.
-
-    Attributes:
-        session (Session): Session object.
-        variant_annotation_path (str): Input variant annotation path.
-        study_locus_path (str): Input study-locus path.
-        variant_index_path (str): Output variant index path.
     """
 
-    session: Session = MISSING
-    variant_annotation_path: str = MISSING
-    credible_set_path: str = MISSING
-    variant_index_path: str = MISSING
-
-    def __post_init__(self: VariantIndexStep) -> None:
-        """Run step."""
+    def __init__(
+        self: VariantIndexStep,
+        session: Session,
+        variant_annotation_path: str,
+        credible_set_path: str,
+        variant_index_path: str,
+    ) -> None:
+        """Run VariantIndex step.
+
+        Args:
+            session (Session): Session object.
+            variant_annotation_path (str): Variant annotation dataset path.
+            credible_set_path (str): Credible set dataset path.
+            variant_index_path (str): Variant index dataset path.
+        """
         # Extract
-        va = VariantAnnotation.from_parquet(self.session, self.variant_annotation_path)
+        va = VariantAnnotation.from_parquet(session, variant_annotation_path)
         credible_set = StudyLocus.from_parquet(
-            self.session, self.credible_set_path, recursiveFileLookup=True
+            session, credible_set_path, recursiveFileLookup=True
         )
 
         # Transform
@@ -42,6 +39,6 @@ def __post_init__(self: VariantIndexStep) -> None:
 
         (
             vi.df.write.partitionBy("chromosome")
-            .mode(self.session.write_mode)
-            .parquet(self.variant_index_path)
+            .mode(session.write_mode)
+            .parquet(variant_index_path)
         )
diff --git a/src/otg/window_based_clumping.py b/src/otg/window_based_clumping.py
index 690bdd882..58e36d15c 100644
--- a/src/otg/window_based_clumping.py
+++ b/src/otg/window_based_clumping.py
@@ -1,56 +1,50 @@
 """Step to run window based clumping on summary statistics datasts."""
 from __future__ import annotations
 
-from dataclasses import dataclass, field
-
-from omegaconf import MISSING
-
 from otg.common.session import Session
 from otg.dataset.summary_statistics import SummaryStatistics
 
 
-@dataclass
 class WindowBasedClumpingStep:
-    """Apply window based clumping on summary statistics datasets.
-
-    Attributes:
-        session (Session): Session object.
-        summary_statistics_input_path (str): Path to the harmonized summary statistics dataset.
-        study_locus_output_path (str): Output path for the resulting study locus dataset.
-        inclusion_list_path (str | None): Path to the inclusion list (list of white-listed study identifier). Optional.
-        locus_collect_distance (int | None): Distance, within which tagging variants are collected around the semi-index. Optional.
-    """
-
-    session: Session = MISSING
-    summary_statistics_input_path: str = MISSING
-    study_locus_output_path: str = MISSING
-    inclusion_list_path: str | None = field(default=None)
-    locus_collect_distance: int | None = field(default=None)
-
-    def __post_init__(self: WindowBasedClumpingStep) -> None:
-        """Run the clumping step."""
+    """Apply window based clumping on summary statistics datasets."""
+
+    def __init__(
+        self,
+        session: Session,
+        summary_statistics_input_path: str,
+        study_locus_output_path: str,
+        inclusion_list_path: str | None = None,
+        locus_collect_distance: int | None = None,
+    ) -> None:
+        """Run window-based clumping step.
+
+        Args:
+            session (Session): Session object.
+            summary_statistics_input_path (str): Path to the harmonized summary statistics dataset.
+            study_locus_output_path (str): Output path for the resulting study locus dataset.
+            inclusion_list_path (str | None): Path to the inclusion list (list of white-listed study identifier). Optional.
+            locus_collect_distance (int | None): Distance, within which tagging variants are collected around the semi-index. Optional.
+        """
         # If inclusion list path is provided, only these studies will be read:
-        if self.inclusion_list_path:
+        if inclusion_list_path:
             study_ids_to_ingest = [
-                f'{self.summary_statistics_input_path}/{row["studyId"]}.parquet'
-                for row in self.session.spark.read.parquet(
-                    self.inclusion_list_path
-                ).collect()
+                f'{summary_statistics_input_path}/{row["studyId"]}.parquet'
+                for row in session.spark.read.parquet(inclusion_list_path).collect()
             ]
         else:
             # If no inclusion list is provided, read all summary stats in folder:
-            study_ids_to_ingest = [self.summary_statistics_input_path]
+            study_ids_to_ingest = [summary_statistics_input_path]
 
         (
             SummaryStatistics.from_parquet(
-                self.session,
+                session,
                 study_ids_to_ingest,
                 recursiveFileLookup=True,
             )
             .coalesce(4000)
             # Applying window based clumping:
-            .window_based_clumping(locus_collect_distance=self.locus_collect_distance)
+            .window_based_clumping(locus_collect_distance=locus_collect_distance)
             # Save resulting study locus dataset:
-            .df.write.mode(self.session.write_mode)
-            .parquet(self.study_locus_output_path)
+            .df.write.mode(session.write_mode)
+            .parquet(study_locus_output_path)
         )
diff --git a/tests/test_cli.py b/tests/test_cli.py
new file mode 100644
index 000000000..9297cd978
--- /dev/null
+++ b/tests/test_cli.py
@@ -0,0 +1,48 @@
+"""Test the command-line interface (CLI)."""
+from unittest.mock import patch
+
+import pytest
+from hydra.errors import ConfigCompositionException
+from omegaconf.errors import MissingMandatoryValue
+from otg.cli import main
+
+
+def test_main_no_step() -> None:
+    """Test the main function of the CLI without a valid step."""
+    override_key = "step"
+    available_steps = [
+        "colocalisation",
+        "eqtl_catalogue",
+        "finngen_studies",
+        "finngen_sumstat_preprocess",
+        "gene_index",
+        "gwas_catalog_ingestion",
+        "gwas_catalog_study_curation",
+        "gwas_catalog_study_inclusion",
+        "gwas_catalog_sumstat_preprocess",
+        "ld_based_clumping",
+        "ld_index",
+        "locus_to_gene",
+        "overlaps",
+        "pics",
+        "ukbiobank",
+        "variant_annotation",
+        "variant_index",
+        "variant_to_gene",
+        "window_based_clumping",
+    ]
+    opts = "\n\t".join(available_steps)
+    expected = f"You must specify '{override_key}', e.g, {override_key}=<OPTION>\nAvailable options:\n\t{opts}"
+
+    with patch("sys.argv", ["cli.py"]), pytest.raises(
+        ConfigCompositionException, match=expected
+    ):
+        main()
+
+
+def test_main_step() -> None:
+    """Test the main function of the CLI complains about mandatory values."""
+    with patch("sys.argv", ["cli.py", "step=gene_index"]), pytest.raises(
+        MissingMandatoryValue, match="Missing mandatory value: step.target_path"
+    ):
+        main()