Expand description
Differential expression analysis for count data.
Provides two methods for identifying differentially expressed genes between two conditions:
- Negative binomial Wald test (
DeMethod::NegativeBinomial) — models count overdispersion (DESeq2-style pipeline with size-factor normalization, method-of-moments dispersion, and Wald z-test). - Wilcoxon rank-sum test (
DeMethod::Wilcoxon) — non-parametric alternative using the existingcrate::testing::mann_whitney_u.
Both methods apply Benjamini-Hochberg correction via
crate::correction::benjamini_hochberg.
The volcano_plot function converts results into points suitable for
plotting (log2 fold-change vs. −log10 adjusted p-value).
Structs§
- DeGene
Result - Per-gene differential expression result.
- DeResults
- Aggregate results from a differential expression analysis.
- Volcano
Point - A point for a volcano plot.
Enums§
- DeMethod
- Which statistical method to use for differential expression.
Functions§
- differential_
expression - Run differential expression analysis on a count matrix.
- volcano_
plot - Convert DE results into volcano-plot points.