Build gene sets to feed gesel

Overview

This repository contains code to build the gene set database for gesel, the client-side gene set search interface.

• genes/build.R will create gene identity files (Ensembl/Entrez identifiers, symbols) for various species, using org.*.eg.db and EnsDb resources from Bioconductor's AnnotationHub.
• go/build.R will extract gene ontology sets from Bioconductor's org.*.eg.db databases.
• msigdb/build.R will extract the MSigDB sets with permissive licenses.
• custom/ contains scripts to prepare GMT files for custom gene sets, e.g., manually added from publications.

All of these gene sets are assembled into Gesel database indices via the top-level build.R script.

Pre-computed database files are available on the Releases page.

Instructions

Interested users can contribute their own gene sets by following these instructions. This is typically used for small collections of gene sets, e.g., manually scraped from publications.

On rare occasions, users might want to suggest the addition of a large pre-assembled collection of gene sets (a la MSigDB). This warrants a new top-level directory with its own build.R script.

Gesel maintainers can trigger a new build of any particular resource by simply updating the relevant VERSION file. This will automatically run the appropriate workflow to create a new release for that resource.

Further links

The specification for the database files is described here.

Applications should use the R, Python or Javascript clients to interface with the Gesel database indices.