Answers are grounded in this paper and cite their sources.
Genome-scale differential expression, CRISPR screen and expression QTL analyses fit millions of count regressions and estimate p-values from billions of random draws. RapidGLM, a C++ library, makes both steps fast, with Newton fits specialized to count models and draws made in batches. Its DESeq2, edgeR, SCEPTRE and cis-QTL analyses agree with the reference tools and run 2.5-17.8 times faster on the same cores. On one compute node, a CRISPR screen's 28.0 million tests took SCEPTRE 11.6 hours and RapidGLM 44 minutes, or 2.8 hours with more accurate small p-values. RapidGLM is MIT-licensed, with C, Python, R and command-line interfaces.
No takes yet. Share an insight, caveat, or question.
Hung et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: