This Snakemake workflow performs the anlysis of long read (Nanopore) repeat sequence data. It calculates statistics and generates plots (custom Python and R scripts) to characterize the nature of the repeat expansion in length and motif composition. Specific alleles (length and motif) can be defined manually for enhanced visualization.
This software is basically a Snakemake workflow that starts with fastq files and runs several known tools in order to analyze reads that cover repeat expansions. It was used in our publication to understand repeat expansions in human FGF14, but it can also be used on other regions of the genome. Long read technology (e.g. Nanopore) is particularly suited for this kind of analysis.
In its core, this workflow – beside doing quality filtering (bbduk) and running some default QC tools – identifies reads with specific regions of interest by their upstream and downstream flanking sequences (also using bbduk). We refer to them as repeat spanning reads. They are the subject to further analyzes steps. The user can specifiy repeat motifs which are visualized in several ways, i.e. the exact locations of the motifs are determined and visualized in graphs (custom scripts). The user can further specify alleles by indicating nucleotide sequences in a yaml file that have to be included or excluded.
git clone https://github.com/kilpert/FGF14_analyses.git
The workflow needs fastq.gz files as input. It was tested on fastq files generated from the FGF14 basecalling workflow, but this is no requirement.
The path to the folder holding the fastq.gz files has to be configured in the config/config.yaml file, e.g.:
fastq_dir: path/to/fastq
The workflow can be started by running the run.sh on the shell
by executing the Snakemake command directly:
snakemake --cores --use-conda --conda-frontend mamba -p --rerun-incomplete
All required software packages will be installed via conda/mamba on the first run.
A demo dataset (demo/input/demo.fastq.gz, 10 reads only) is provided. It can be used for testing the workflow. The run time is about 15s.
Run the workflow on the demo data:
snakemake --cores --use-conda --conda-frontend mamba -p --rerun-incomplete --configfile config/demo.config.yaml
The demo output will be saved to the demo/output folder as specified in the demo/demo.config.yaml.
The exact demo output is already provided in the demo/output.tar.gz file. Unzip to restore the original output folder from the workflow (e.g. for comparison).
The workflow was tesed on:
- Ubuntu 22.04.2 LTS
- conda 23.3.1
- mamba 1.4.2
- biopython 1.81
- Snakemake 7.32.4
Mohren et al. (2024). Advancing molecular, phenotypic and mechanistic insights of FGF14 pathogenic expansions (SCA27B). in prep.