This repository contains all code and scripts used to perform the analyses described in the paper "Pathogenic variants in actionable MODY genes associate with type 2 diabetes" (DOI:10.1038/s42255-020-00294-3).
The repository includes both bioinformatics and biostatistics analyses.
Note: the code has not been tested outside of its original environment and depends on data not publicly available.
The bioinformatics analysis was performed using the scripts in the bioinformatics directory.
The different analysis steps are listed in the plain text file bioinformatics/template_analysis, which is provided to the bioinformatics/start_multiserveur.pl script.
Then the bioinformatics/start_sample.pl script is launched with all the different steps of the analysis.
All the different scripts to perform each steps can be found in the bioinformatics/Include directory.
These scripts use local databases:
- MySQL databases with informations about the different samples (e.g., sample ID, sex, demultiplexing index and lane, raw data path) and the dbSNP annotations (version 135).
- A MongoDB database with the dbNSFP annotations (version 3.0).
- A Redis database with mutation counting informations.
Tools, captures and sequencing adaptors paths can be passed by the files in the bioinformatics/Config directory.
Perl version used is Perl 5, version 16, subversion 3 (v5.16.3) on CentOS Linux 7 (Core, x86_64).
The statistical analyses were performed using the scripts in the biostatistics directory.
All scripts are numbered according to the order of use.
The directory biostatistics/utils contains additional scripts needed in the analyses.
The R version used is the 3.6.3 (2020-02-29) on Debian GNU/Linux 9 (stretch, x86_64) available as a Docker image (umr1283/stat:R363).
For questions and other discussion, please contact the authors.