Bioinformatics support as code

This project contains functionality for describing bioinformatic analyses as code.

Quick start

Install the dependencies:

$ pip install -r requirements.txt
$ cd smarttools/smarttoolbase && python setup.py install && cd ../..

Create the file analysis.yml with the content below.

input_dataset_uri: runners/example_analysis/data/todo_lists
output_dataset_base: /tmp
output_dataset_name_suffix: top_4_things
local_smarttool_fpath: smarttools/smarttools/simple_example_tool/smarttool_runner.py

Run the analysis:

$ python runners/scripts/local_runner.py analysis.yml
Created: file:///tmp/todo_lists_top_4_things

Repository overview

The directory smarttools contains code and Docker image definitions for building so called Smart tools.

A Smart tool is a docker/singularity containerised analysis that takes standardised input and output arguments, which allows it to be run by something else that understands those arguments. This means that it is possible to write other tools, so called "runners", that can process all the items in the input dataset.

These tools runners are present in the runners directory. Smarttool runners are programs designed to take input Dtool datasets and smartools and produce the scripts or directly execute commands required to process the dataset.

Some Smart tools require the input dataset to have specific overlays specified. For example the align_seqs_hisat2 Smart tool requires the overlays is_read1 and pair_id to be set. These can be generated using overlay creator scripts stored in the directory overlay_creators. To create the is_read1 and pair_id overlays on a dataset with fastq files one could use the create_paired_read_overlays_from_fname.py script.