Analysis code MMSplice manuscript

MMSplice predict variant effect on PSI, PSI5, PSI3, splicing efficiency and pathogenicity. This repository contains analysis code for MMSplice manuscript. MMSplice version mmsplice==0.2.0.

Software dependencies

The analysis was done with python and ploting was done with R. Required packages are listed in env.yaml.

1. Install miniconda or anaconda.
2. Run: conda env create -f env.yaml. This will install a new conda environment mmaplice-manuscript.
3. Activate the environment: source activate mmaplice-manuscript. Run the analysis code (in jupyter-notebooks) in this environment.
4. Install COSSMO from https://github.com/PSI-Lab/COSSMO

Folder structure

In general, the jupyter-notebooks were used to process, analysis the data and perform predictions. The result table are saved and read by R scripts for plotting.

• script/Figure2: Code to generate figure for the Vex-seq and MFASS data.
• script/Figure3: Code to generate figure for GTEx A5SS amd A3SS variants.
• script/Figure4: Code to generate figure for MaPSy data.
• script/Figure5: Code to generate figure for ClinVar data. Result table of other models were obtained from [Kipoi manuscript] (https://github.com/kipoi/manuscript)
• data/: contains raw and processed data