Welcome to lcr-modules’s Documentation!¶
Getting Started With lcr-modules¶
- Users: Check out this Getting Started guide and the Demo Project.
- Contributors: Check out this Getting Started guide and the lcr-modules repository.
Installing oncopipe¶
Due to active development of oncopipe, you should always install the development version of oncopipe
from the lcr-modules repository. This way, your installation will automatically update when you pull from GitHub.
git clone https://github.com/LCR-BCCRC/lcr-modules.git
pip install -e lcr-modules/oncopipe
Motivation¶
This project aims to become a collection of standard analytical modules for genomic and transcriptomic data. Too often do we copy-paste from each other’s pipelines, which has several pitfalls:
* Too much time spent on routine analyses * Increased risk for hidden logical bugs
* Duplicated effort within and between labs * No consistently used pipelining tool
* Inefficient dissemination of best practices * Steep learning curve for new members
Fortunately, all of these problems can be solved with standardized analytical modules, and the benefits are many:
* Projects can ramp up faster * Consistent intermediate/output files
* Streamline efforts between labs * More reproducible analyses
* Define analytical best practices * Easier-to-write methods
* Consolidate collective expertise * Automated logging and “paper trail”
* Simplify member onboarding * Easier peer review of code
* And happier bioinformaticians!
What Are Modules?¶
Each module accomplishes a specific analysis, generally centered around a specific tool (e.g. Strelka2, Manta, MutSigCV). Analyses—and by extension, modules—can be organized into different levels. The figure below contains for examples for each level.
- Level-1 Analyses: They process raw sequencing data, generally producing BAM/FASTQ files.
- Level-2 Analyses: They perform sample-level analyses on level-1 output, such as variant calling and gene expression quantification.
- Level-3 Analyses: They aggregate sample-specific level-2 output and perform cohort-wide analyses, such as the identification of sifgnificantly mutated genes.
- Level-4 Analyses: They are project-specific and are meant to ask specific questions of the data. These are the analyses you ideally want to spend your time on.