genio Gen I/O

The genio (GENetics I/O) package provides easy-to-use and efficient readers and writers for formats in genetics research. Currently targets Plink, Eigenstrat, and GCTA formats (more to come). Plink BED/BIM/FAM and GCTA GRM formats are fully supported. Lightning fast read_bed and write_bed (written in Rcpp) reads and writes genotypes between native R matrices and Plink BED format. make_* functions create default FAM and BIM files to go with simulated genotype data. Otherwise, the package consists of wrappers for readr functions that add missing extensions and column names (often absent in these files).

Installation

You can install the released version of genio from CRAN with:

install.packages("genio")

Install the latest development version from GitHub:

install.packages("devtools") # if needed
library(devtools)
install_github("OchoaLab/genio", build_vignettes = TRUE)

You can see the package vignette, which has more detailed documentation, by typing this into your R session:

vignette('genio')

Example

Load library:

library(genio)

Make a BED/BIM/FAM file set for simulated data

Note that write_plink writes all three BED/BIM/FAM files together, while each write_{bed,bim,fam} function creates a single file.

# write your genotype matrix stored in an R native matrix

# (here we create a small example with random data)
# create 10 random genotypes
X <- rbinom(10, 2, 0.5)
# replace 3 random genotypes with missing values
X[sample(10, 3)] <- NA
# turn into 5x2 matrix
X <- matrix(X, nrow = 5, ncol = 2)

# also create a simulated phenotype vector
pheno <- rnorm(2) # two individuals as above

# write simulated data to all BED/BIM/FAM files in one handy command
# missing BIM and FAM columns are automatically generated
# data dimensions are validated for provided data
write_plink('random', X, pheno = pheno)

### same thing in separate steps:

# create default tables to go with simulated genotype data
fam <- make_fam(n = 2)
bim <- make_bim(n = 5)
# overwrite with simulated phenotype
fam$pheno <- pheno

# write simulated data to BED/BIM/FAM separately (one command each)
# extension can be omitted and it still works!
write_bed('random', X)
write_fam('random', fam)
write_bim('random', bim)

Reading and writing existing data

# read individual and locus data into "tibbles"

# read plink data all at once
data <- read_plink('sample')
# extract genotypes and annotation tables
X   <- data$X
bim <- data$bim
fam <- data$fam

# Plink files read individually
bim <- read_bim('sample.bim')
fam <- read_fam('sample.fam')
X   <- read_bed('sample.bed', nrow(bim), nrow(fam))

# Eigenstrat formats
snp <- read_snp('sample.snp')
ind <- read_ind('sample.ind')

# in all cases extension can be omitted and it still works!
bim <- read_bim('sample')
fam <- read_fam('sample')
snp <- read_snp('sample')
ind <- read_ind('sample')

# write these data to other files
# here extensions are also added automatically
# write all plink files together, ensuring consistency
write_plink('new', X, bim, fam)
# write plink files individually
write_fam('new', fam)
write_bim('new', bim)
write_bed('new', X)
# Eigenstrat files
write_ind('new', ind)
write_snp('new', snp)

Reading and writing GCTA GRM files

# read data from GRM files:
# - sample.grm.bin (kinship matrix),
# - sample.grm.N.bin (sample sizes matrix), and
# - sample.grm.id (family and ID table for individuals in this data)
obj <- read_grm( 'sample' )
# the kinship matrix
kinship <- obj$kinship
# the pair sample sizes matrix
M <- obj$M
# the fam and ID tibble
fam <- obj$fam

# write data into new GRM files
# writes: new.grm.bin, new.grm.N.bin, new.grm.id
write_grm( 'new', kinship, M = M, fam = fam )