CRAN Task View: Statistical Genetics

Great advances have been made in the field of genetic analysis over the last years. The availability of millions of single nucleotide polymorphisms (SNPs) in widely available databases, coupled with major advances in SNP genotyping technology that reduce costs and increase throughput, are enabling a host of studies aimed at elucidating the genetic basis of complex disease. The focus in this task view is on R packages implementing statistical methods and algorithms for the analysis of genetic data and for related population genetics studies.

A number of R packages are already available and many more are most likely to be developed in the near future. Please send your comments and suggestions to either of the task view maintainer's Giovanni Montana or Neil Shephard .

• Population Genetics : genetics implements classes and methods for representing genotype and haplotype data, and has several functions for population genetic analysis (e.g. functions for estimation and testing of Hardy-Weinberg and linkage disequilibria, etc.). Geneland has functions for detecting spatial structures from genetic data within a Bayesian framework via MCMC estimation. rmetasim provides an interface to the metasim engine for population genetics simulations. hapsim simulates haplotype data with pre-specified allele frequencies and LD patterns. A few population genetics functions are also implemeted in gap and popgen. popgen has functions for clustering SNP genotype data and SNP simulation from a Multinomial-Dirichlet model. hierfstat allows the estimation of hierarchical F-statistics from haploid or diploid genetic data. LDheatmap creates a heat map plot of measures of pairwise LD. mapLD measures linkage disequilibrium and constructs haplotype blocks. hwde fits models for genotypic disequilibria. Whilst HardyWeinberg provides graphical representation of disequilibria via ternary plots (also known as de Finetti diagrams). Biodem package provides functions for Biodemographical analysis, e.g. Fst() calculates the Fst from the conditional kinship matrix. Package kinship offers some functions for analysis on pedigrees. The adegenet implements a number of different methods for analysing population structure using multivariate statistics, graphics and spatial statistics.
• Phylogenetics : Phylogenetic and evolution analyses can be performed via ape and apTreeshape. Package ouch provides Ornstein-Uhlenbeck models for phylogenetic comparative hypotheses. PHYLOGR is a suite of functions for the analysis of phylogenetically simulated data sets and phylogenetically-based GLS model fitting. stepwise implements a method for stepwise detection of recombination breakpoints in sequence alignments.
• Linkage and Association : Packages in this category provide statistical methods to test associations between genetic markers and a phenotype. gap is a package for genetic data analysis of both population and family data; it contains functions for sample size calculations, probability of familial disease aggregation, kinship calculation, and some tests for linkage and association analyses. Among the other functions, genecounting() estimates haplotype frequencies from genotype data, and gcontrol() implements a Bayesian genomic control statistics for association studies. For family data, tdthap offers an implementation of the Transmission/Disequilibrium Test (TDT) for extended marker haplotypes, whereas powerpkg performs power analyses for the affected sib pair and the TDT design. The catmap can be used for the meta-analysis of case-contorl and TDT data.
• Linkage Disequilibrium and haplotype mapping : The package hapassoc performs likelihood inference of trait associations with haplotypes in GLMs, and haplo.ccs estimates haplotype and covariate relative risks in case-control data by weighted logistic regression. haplo.stats also contains functions for the analysis of indirectly measured haplotypes. The statistical methods assume that all subjects are unrelated and that haplotypes are ambiguous (due to unknown linkage phase of the genetic markers). SimHap is another package that infers haplotypes and tests for associations and incorporates multiple-imputation to deal with haplotype phase uncertainty. tdthap implements transmission/disequilibrium tests for extended marker haplotypes. ldDesign is a package for design of experiments for association studies for detection of linkage disequilibrium. LDheatmap creates a heat map plot of measures of pairwise LD. mapLD measures linkage disequilibrium and constructs haplotype blocks.
• Genome-Wide Association Studies (GWAS) : With recent technical advances in high-throughput genotyping technologies the possibility of performing Genome-Wide Association Studies is now a feasible strategy. A number of packages are available to facilitate the analysis of these large data sets. GenAbel is designed for the efficent storage and handling of GWAS data with fast analysis tools for quality control, association with binary and quantitative traits, as well as tools for visualizing results. pbatR provides a GUI to the powerful PBAT software which performs family and population based family and population based studies. The software has been implemented to take advantage of parallel processing, which vastly reduces the computational time required for GWAS. SNPassoc is another package for carrying out GWAS analysis. It provides descriptive statistics of the data (inlcuding patterns of missing data) and tests for Hardy-Weinberg equilibrium. Single-point analyses with binary or quantitative traits are implemented via generalized linear models, and multiple SNPs can be anlaysed for haplotypic associations or epistasis. snpMatrix Implements classes and methods for large-scale SNP association studies.
• QTL mapping : Packages in this category develop methods for the analysis of experimental crosses to identify markers contributing to variation in quantitative traits. bqtl implement both likelihood-based and Bayesian methods for inbred crosses and recombinant inbred lines. qtl provides several functions and a data structure for QTL mapping, including a function scanone() for genome-wide scans. The package qtlDesign has functions for designing QTL experiments, including power computations. qtlbim implements Bayesian Interval Mapping for QTL.
• Multiple testing : The package qvalue implements False Discovery Rate; the main function qvalue() estimates the q-values from a list of p-values. Package multtest also offers several non-parametric bootstrap and permutation resampling-based multiple testing procedures.
• Importing Sequence Data : There are utilities in the seqinr package to import sequence data from various sources, including files of aligned sequences in mase, clustal, phylip, fasta and msf format which will be of utility to some population genetic analysis. Users interested in using R for sequence data and bioinformatics are also referred to the BioConductor project.

CRAN packages:

• adegenet
• ape
• apTreeshape
• Biodem
• bqtl
• catmap
• gap (core)
• GenABEL
• Geneland
• genetics (core)
• hapassoc
• haplo.ccs
• haplo.stats (core)
• hapsim
• HardyWeinberg
• hierfstat
• hwde
• kinship
• ldDesign
• LDheatmap
• lodplot
• luca
• mapLD
• multtest
• ouch
• pbatR
• PHYLOGR
• popgen
• powerpkg
• qtl
• qtlDesign
• qvalue
• rmetasim
• seqinr
• SimHap
• SNPassoc
• SNPmaxsel
• snp.plotter
• stepwise
• tdthap
• untb

Related links:

• The Rgenetics Project
• R Software by David Clayton
• BayesMendel, R software for predicting who may carry a cancer susceptibility gene
• R genetics project
• R code for estimating haplotype frequencies from pooled DNA samples
• Computer programs by Jing Hua Zhao (some are included in the gap package)
• An alphabetic list of genetic analysis software
• Hclust.R, software for Choosing Tag SNPs
• Bayesian Adaptive Regression Splines (BARS)
• Direct Simulation Approach (DAS), for simulation of p-values
• Vector Generalized Linear and Additive Models (VGLMs/VGAMs)
• P2BAT: Tools for the statistical analysis of family-based association studies.
• R Software by David Duffy
• MareyMap: GUI for comparing physical and genetic maps for estimation of recombination rates.
• p_ACT : R program that adjusts sets of up to 1000 p-values from association tests between correlated traits and SNPs for multiple testing, accounting for the correlation between tests.
• BioConductor: an open source and open development software project for the analysis and comprehension of genomic data based on R.