| crlmm {oligo} | R Documentation |
Performs genotype calls via CRLMM (Corrected Robust Linear Model with Maximum-likelihood based distances).
crlmm(object, correction = NULL, recalibrate = TRUE, minLLRforCalls = c(5, 1, 5), verbose = TRUE, correctionFile = NULL, prefix = "tmp.crlmm.", balance = 1.5) justCRLMM(filenames, batch_size = 40000, minLLRforCalls = c(5, 1, 5), recalibrate = TRUE, balance = 1.5, phenoData = NULL, verbose = TRUE, pkgname = NULL)
object |
SnpQSet object |
filenames |
character vector with the filenames. |
batch_size |
integer defining how many SNPs should be processed at a time. |
correction |
The output of the EM algorithm |
recalibrate |
Logical - should recalibration be performed? |
prefix |
String defining the prefix to be used with the temporary files. |
balance |
Control parameter to balance homozygotes and heterozygotes calls. |
minLLRforCalls |
Minimum thresholds for genotype calls. |
verbose |
Logical. |
correctionFile |
A filename. |
phenoData |
phenoData object or NULL |
pkgname |
alt. pdInfo package to be used |
The correctionFile is a string (eg, "outputEM.rda")
pointing to a filename. If the file does not exist, crlmm will
save a file with that name containing the results of the EM
algorithm. If the file exists, its content is loaded and used by
CRLMM. The correctionFile is meant to save time if crlmm
is be run multiple times, as the EM algorithm does not need to be run
everytime.
The justCRLMM method is more efficient in terms of memory. It
uses the CEL files directly, taking SNPs by batch.
SnpCallSetPlus object.
## crlmmResults <- justCRLMM(list.celfiles())