MutSig2CV

MutSig2CV identifies genes significantly mutated in cancer by analyzing somatic point mutations in exome sequences to detect genes mutated more often than expected from background mutation processes.

Key Features:

• Significant mutation identification: Evaluates lists of somatic point mutations from DNA sequencing to pinpoint genes with mutation frequencies above background expectations.
• Driver versus passenger distinction: Differentiates likely driver mutations from passenger mutations that are not causally linked to cancer progression.
• Mutational heterogeneity correction: Incorporates an analytical methodology that accounts for heterogeneity across samples to reduce false positives.
• False-positive reduction: Reduces implausible gene calls such as olfactory receptors and large muscle proteins like titin that arise from artefactual signals.
• Large-cohort exome analysis: Demonstrated on datasets including analyses of 4,742 human cancers across 21 types and on exome sequences from ~3,083 tumor-normal pairs.
• Insights into mutational processes: Reveals variation in mutation frequency and spectrum within cancer types and identifies correlations with DNA replication timing and transcriptional activity.
• Sample size guidance: Provides estimates suggesting that near-saturation for clinically important mutations may require sample sizes of roughly 600 to 5,000 per tumor type.

Scientific Applications:

• Cancer gene cataloguing: Produces a more accurate and comprehensive catalogue of significantly mutated genes across tumor types.
• Prioritization for follow-up studies: Ranks candidate genes for experimental validation by distinguishing likely drivers from artefactual and passenger mutations.
• Study of mutational etiology: Enables analysis of mutational heterogeneity and correlations with replication timing and transcriptional activity to inform biological hypotheses.
• Experimental design and power estimation: Informs sample size requirements for discovering recurrent, clinically relevant mutations in specific tumor types.

Methodology:

Evaluates somatic point mutation lists from exome sequences of tumor-normal pairs, compares per-gene observed mutation frequencies to expectations from background mutation processes, incorporates heterogeneity-aware analytical methods, and analyzes mutation frequency and spectrum with correlations to DNA replication timing and transcriptional activity.