ourMELONS/R/greedyMix.R

108 lines
4.1 KiB
R
Raw Normal View History

2020-05-20 15:34:40 +02:00
#' @title Clustering of individuals
#' @param data data file
#' @param format Data format. Format supported: "FASTA", "VCF" ,"BAM", "GenePop"
#' @param partitionCompare a list of partitions to compare
#' @param npops number of populations
#' @param counts counts
#' @param sumcounts sumcounts
#' @param max_iter maximum number of iterations
#' @param alleleCodes allele codes
#' @param inp input file
#' @param popnames population names
#' @param fixedK if \code{TRUE}, the number of populations is fixed
#' @param verbose if \code{TRUE}, prints extra output information
2020-06-24 11:48:23 +02:00
#' @importFrom utils read.delim
2021-09-03 11:17:00 +02:00
#' @importFrom vcfR read.vcfR
2021-09-03 12:56:00 +02:00
#' @importFrom Rsamtools scanBam
#' @importFrom adegenet read.genepop .readExt
2021-09-03 12:50:11 +02:00
#' @references Samtools: a suite of programs for interacting
#' with high-throughput sequencing data. <http://www.htslib.org/>
2020-05-20 15:34:40 +02:00
#' @export
#' @examples
2024-08-07 11:42:29 +02:00
#' data <- system.file("extdata", "BAPS_clustering_diploid.txt", package = "rBAPS")
2024-04-08 13:29:03 +02:00
#' greedyMix(data, "baps")
greedyMix <- function(
2024-04-10 14:25:45 +02:00
data, format = gsub("^.*\\.", "", data), partitionCompare = NULL, npops = 3L,
counts = NULL, sumcounts = NULL, max_iter = 100L, alleleCodes = NULL,
inp = NULL, popnames = NULL, fixedK = FALSE, verbose = FALSE
) {
# Importing and handling data ================================================
# TODO: use format as class and make handling data a generic
if (tolower(format) %in% "fasta") {
2024-08-09 14:57:07 +02:00
data <- convert_FASTA_to_BAPS(data)
format <- "baps"
}
2024-03-25 16:05:00 +01:00
if (tolower(format) %in% "baps") {
data <- process_BAPS_data(data, NULL)
c <- list(
noalle = data[["noalle"]],
data = data[["data"]],
adjprior = data[["adjprior"]],
priorTerm = data[["priorTerm"]],
rowsFromInd = data[["rowsFromInd"]],
Z = data[["Z"]],
dist = data[["dist"]]
)
} else if (tolower(format) %in% "genepop") {
data <- process_GenePop_data(data)
c <- list(
noalle = data[["noalle"]],
data = data[["data"]],
adjprior = data[["adjprior"]],
priorTerm = data[["priorTerm"]],
rowsFromInd = data[["rowsFromInd"]],
Z = data[["Z"]],
dist = data[["dist"]]
)
2024-03-25 16:05:00 +01:00
} else {
data <- importFile(data, format, verbose)
data <- handleData(data, tolower(format))
c <- list(
noalle = data[["noalle"]],
data = data[["newData"]],
adjprior = data[["adjprior"]],
priorTerm = data[["priorTerm"]],
rowsFromInd = data[["rowsFromInd"]],
Z = data[["Z"]],
dist = data[["dist"]]
)
}
# Comparing partitions =======================================================
2024-04-08 10:11:25 +02:00
ninds <- length(unique(c[["data"]][, ncol(c[["data"]])]))
if (!is.null(partitionCompare)) {
logmls <- comparePartitions(
c[["data"]], nrow(c[["data"]]), partitionCompare[["partitions"]], ninds,
c[["rowsFromInd"]], c[["noalle"]], c[["adjprior"]]
)
}
# Generating partition summary ===============================================
ekat <- seq(1L, ninds * c[["rowsFromInd"]], c[["rowsFromInd"]])
c[["rows"]] <- cbind(ekat, ekat + c[["rowsFromInd"]] - 1L)
2024-09-27 06:59:18 +02:00
logml_npops_partitionSummary <- indMixWrapper(c, npops, counts, sumcounts, max_iter, fixedK, verbose) # FIXME: not working for FASTA
logml <- logml_npops_partitionSummary[["logml"]]
npops <- logml_npops_partitionSummary[["npops"]]
partitionSummary <- logml_npops_partitionSummary[["partitionSummary"]]
# Generating output object ===================================================
out <- list(
2023-08-11 14:31:08 +02:00
"alleleCodes" = alleleCodes, "adjprior" = c[["adjprior"]],
"popnames" = popnames, "rowsFromInd" = c[["rowsFromInd"]],
"data" = c[["data"]], "npops" = npops, "noalle" = c[["noalle"]],
"mixtureType" = "mix", "logml" = logml
)
if (logml == 1) {
return(out)
}
# Writing mixture info =======================================================
changesInLogml <- writeMixtureInfo(
logml, c[["rowsFromInd"]], c[["data"]], c[["adjprior"]], c[["priorTerm"]],
NULL, inp, partitionSummary, popnames, fixedK, verbose
)
# Updateing results ==========================================================
2023-09-14 11:02:06 +02:00
return(c(out, list("changesInLogml" = changesInLogml)))
}