The codes are used for processing deep-targeted UMI-seq data.
The workflows are written in gwf.
It includes the steps of both UMI processing and variant calling options of
shearwater(both AND & OR algorithm),
We used it for benchmarking different variant callers and UMI processing strategies for detecting low-frequency mutations from cfDNA colorectal cancer patients. For detailed theories behind each method, and the specific step for modification based on your own use, please also check the related github respectively.
Directory workflow includes all steps for processing cancer patient plasma sample, from UMI processing to variant calling. Please check the README under it for more information.
For running GWF workflow, please locate at the working directory which includes both the templates.py and workflow.py.
Specifically, the steps under workflow, can be performed by
cd workflow
gwf -b slurm run
Remember to change the account, running time, core and memory for the functions inside templates.py for your own use.
Directory workflow_PON includes all steps for processing PON which is required by variant calling of mutect2, shearwater and DREAMS-vc. Therefore, before performing variant calling for cancer plasma sample, please check the README under this directory and follow all steps to prepare all files in need.
For running GWF workflow, please locate at the working directory which includes both the templates.py and workflow.py.
Specifically, the steps under workflow_PON, can be performed by
cd workflow_pon
gwf -b slurm run
Remember to change the account, running time, core and memory for the functions inside templates.py for your own use.
As shown by the json file under workflow and workflow_PON, we utilized three different conda environments to run the workflow.
- For conda environment named 'umiseq', please run
conda create -n umiseq -c bioconda -c gwforg python=3.7 gwf samtools bwa picard umi_tools fgbio gatk4 varscan pysamstats seqtk
please check environment_umiseq.yml for all dependencies and related packages.
- For conda environment named 'shearwater', please run
conda create -n shearwater
conda activate shearwater
conda install conda-forge::r-tidyverse
To install package 'deepSNV', start R (version "4.4") and enter:
if (!require("BiocManager", quietly = TRUE))
install.packages("BiocManager")
BiocManager::install("deepSNV")
please check environment_sw.yml for all dependencies and related packages.
- For conda environment named 'dreams_M', please run
conda create -n dreams_M
conda activate dreams_M
Under R, install DREAMS package by
install.packages("devtools")
devtools::install_github("JakobSkouPedersenLab/dreams")
please check environment_dreams.yml for all dependencies and related packages.