Run Beagle4.1 to impute a target VCF with a reference VCF

Beagle4.1 is slower than Beagle5 by far. However, it can use genotype-likelihoods (the GL VCF field) for its algorithm, which is potentially more accurate. In cases like imputing GBS data, or if you have time to wait around, this should be better than Beagle5.0 for a first pass imputation of observed sites. Impute with Beagle V5.0. Use an "imputation reference panel". Refer to Beagle documentation for meaning of arguments passed.

runBeagle4pt1GL(
  targetVCF,
  refVCF,
  mapFile,
  outName,
  nthreads,
  maxmem = "500g",
  impute = TRUE,
  ne = 1e+05,
  samplesToExclude = NULL,
  niter = 10
)

Arguments

targetVCF

passes to Beagle `gl=targetVCF`

refVCF

passes to Beagle `ref=targetVCF`

mapFile

passes to Beagle `map=mapFile`

outName

passes to Beagle `out=outName (don't put file suffix, Beagle adds *.vcf.gz).

nthreads

passes to Beagle `nthreads=nthreads`

maxmem

passes to java `-Xmx<maxmem>`

impute

passes to Beagle `impute=TRUE`

ne

passes to Beagle `ne=ne`

niter

passes to Beagle `niterations=niter`

Value

Details

NOTICE: This function is part of a family of functions ("imputation_functions") developed as part of the NextGen Cassava Breeding Project genomic selection pipeline. For some examples of their useage:

• IITA_2021GS (should be first use of this package)

• NRCRI_2021GS

• TARI_2020GS

See also

Other imputation_functions: convertDart2vcf(), convertVCFtoDosage(), createGenomewideDosage(), filter_positions(), mergeVCFs(), postImputeFilterBeagle4pt1(), postImputeFilter(), runBeagle5(), splitVCFbyChr()