RVs-r

RVs-r infers disease risk genes by computing probabilities of rare variant sharing among related individuals in multiplex pedigrees.

Key Features:

• Association and Linkage Testing: Implements tests for both association and linkage between rare variant genotypes and dichotomous phenotypes (e.g., disease status) within family samples using probabilities of rare variant sharing under the null hypothesis.
• Gene-Based Analyses: Performs joint analysis of multiple rare variants within a region (e.g., genes) to increase statistical power compared to single-variant tests.
• Partial and Complete Sharing Tests: Provides partial sharing tests based on probabilities of rare variant sharing among subsets of affected relatives as well as complete sharing tests to address phenocopies.
• Haplotype-Based Variant Definition: Defines rare variants using haplotype information to refine variant definitions and interpret sharing patterns.
• No Requirement for External Frequency Estimates or Control Samples: Conducts analyses without relying on external estimates of variant frequency or separate control samples.
• Assessment and Correction of Assumptions: Includes functionality to assess and address potential violations of key model assumptions underlying the analyses.

Scientific Applications:

• Simulations in multiplex cleft families: Simulations showed that partial and complete sharing versions achieved comparable statistical power to RareIBD and the Gene-Based Segregation Test and superior power relative to the pVAAST linkage statistic.
• Exome analysis of affected relatives: Analysis of rare single nucleotide variants in the exomes of 151 affected relatives from 54 families revealed distinct variant-sharing patterns and no single gene with significant excess sharing, with differentiation between complete and partial sharing tests informing interpretation.

Methodology:

Computes probabilities of rare variant sharing among relatives and uses them to perform association and linkage tests (complete and partial sharing), gene-based joint analyses, and haplotype-based variant definition, with functionality to assess and correct violations of underlying assumptions and without using external frequency estimates or control samples.