Identify Differential Selection

Tests for context-dependent selection on cancer driver genes using somatic mutation data. The package implements the DiffDriver statistical framework to assess whether the strength of positive selection on mutations in a driver gene is associated with tumor- or individual-level context variables, such as clinical traits, genomic features, or immune microenvironment subtypes. DiffDriver estimates individual- and position-specific background mutation rates, models selection as a deviation from the background rate using functional annotations, and tests context effects through a latent-variable logistic model. It provides utilities for preparing mutation and annotation data, fitting differential-selection models, running gene-level association tests, summarizing candidate genes, and visualizing mutation patterns. The method is described in Zhou et al. (2026) "Detecting context-dependent selection on cancer driver genes with DiffDriver" .

diffDriver

To install

First, install the package remotes.

install.packages("remotes")

Then install diffdriver.

remotes::install_github("szhaolab/diffdriver", ref ="main") 

The package also relies on annotation files. There are two types of annotation files.

• 9-annotation files.
• 96-annotation files.

Both sets of files can be downloaded at

How to use

Please checkout the tutorial on the package website.