File Fomat
Ronnie Nelson
The triM format is used in the example files, as well as one phenotype file, and the following format needs to be used. Note that all the files are tab delimited.
The marker information files:
The marker information files provides information on the number of chromosomes, the total number of markers. The following lines contain the total number of markers per chromosome followed by a numeric ID that indicates which column pair in the genotype file contains the genotypes for that marker. This information is given on one line for each chromosome. The following lines contain the space between each marker for each chromosome provided (i.e. the map data), also one line for each chromosome. The last block of lines contains the name of each marker, one line for each chromosome. Note that the first column of each row containing the positions and names of the markers should be filled with a '1'.
The example marker information file: “mrkinfo_test.txt ”
The genotype file:
The genotype file provides the genotypic information for each individual. The individual IDs are indicated in the first column of each row. The following columns are filled with integer values indicating the genotype at each locus in sequence. Every 2 columns correspond to one marker (i.e. one allele in each column) and the columns are arranged sequentially in the order described in the marker file.
The example marker information files: “marker_test.txt”
The pedigree file:
The pedigree file is arranged in full-sib families. For each family, the number of F2 individuals within the family are provided. This is followed by the individual ID's in the first column (starting with the F0 generation, then the F1s and then the F2s). For each individual his/her parent's ID's are provided in the next two columns followed by it's sex. Note, for the F0 generation the parents are indicated with a '0' while an additional column with line origins is provided.
The example pedigree file: “ped_test.txt”
The phenotype file:
The phenotype files provide the information of the phenotypes of all the individuals. The first line contains the heading “ID” in the first column, followed by the phenotype names. The columns are filled with, first the individual ID’s and then the phenotypic values as indicated by the headings.
The example pedigree file: “pheno_test.txt ”
Federico Calboli
Mats
Hi Federico!
The MAPfastR data object is a regula R list, so what you need to do is re-shape and split your current data frame to fit the structure, and you should be good to go. I attach some code that does that for the example you provided below. You should be able to use it on real data with only very minor modifications, I think. The comments describe where some things must be added.
/Mats
#Assume that the data is currently in the data frame "Fed_data"
#In case you want to start from a datafile of the format in the post
#Fed_data <- read.delim("/path/datafile.txt", sep = "", stringsAsFactors = F)
#First we extract and reshape the genotypes
Fed_temp_geno <- t(Fed_data[,5:dim(Fed_data)[2]])
Fed_geno <- as.data.frame(matrix(ncol = dim(Fed_temp_geno)[2]*2, nrow = dim(Fed_temp_geno)[1]/2))
odd <- seq(from= 1, to = dim(Fed_temp_geno)[1], by = 2)
even <- odd + 1
geno_col <- 1
for(col in 1:dim(Fed_temp_geno)[2]){
Fed_geno[, geno_col] <- Fed_temp_geno[odd, col]
Fed_geno[, geno_col+1] <- Fed_temp_geno[even, col]
geno_col <- geno_col + 2
}
#This dataframe you need to construct from your marker information data
Fed_placeholder <- data.frame(chr = 1, sex_1_cM = c(1,2,3), sex_2_cM = c(1,2,3), ref_cM = c(1,2,3))
#Here we finalise the genotypes into MAPfastR from
names(Fed_geno) <- rep(Fed_data[,"id"], each = 2)
Fed_geno <- cbind(Fed_geno, Fed_placeholder)
#Now, the pheno data.frame
#First. we deduce the generation of each individual from the pedigree
Fed_generation <- array(dim=dim(Fed_data)[1])
gen <- 1
prev_gen <- array()
curr_gen <- array()
while (gen <= 3){
hit <- 1
for (ind in 1:dim(Fed_data)[1]){
if (gen == 1){
if (Fed_data[ind, 2] == 0 & Fed_data[ind, 3] == 0){
Fed_generation[ind] <- gen
curr_gen[hit] <- Fed_data[ind, "id"]
hit <- hit + 1
}
}
else{
if (Fed_data[ind, 2] %in% prev_gen & Fed_data[ind, 3] %in% prev_gen){
Fed_generation[ind] <- gen
curr_gen[hit] <- Fed_data[ind, "id"]
hit <- hit + 1
}
}
}
prev_gen <- curr_gen
curr_gen <- array()
gen <- gen + 1
}
#Here you need to add line origin (for the founders in particular)
Fed_line <- array(dim = 10)
#Then we extract, combine and reshuffle
Fed_pheno <- cbind(Fed_generation,Fed_data[,c(4,2,3)], Fed_line)
#We conform to MAPfastR conventions
names(Fed_pheno) <- c("generation", "sex", "parent_1", "parent_2", "line")
Fed_pheno[,"sex"] <- Fed_pheno[,"sex"] + 1
Fed_pheno[,"sex"] <- 2/Fed_pheno[,"sex"]
#Finally, we make a MAPFastR object, and add some bookkeeping parameters
#note the "$heterogam" and make sure that lines up with your sex encoding
Fed_MAPfastR_data <- list(pheno = Fed_pheno, geno = Fed_geno)
Fed_MAPfastR_data$backcross <- 0
Fed_MAPfastR_data$backcross.line <- NA
Fed_MAPfastR_data$backcross.parent <- NA
Fed_MAPfastR_data$sex.restrict <- 0
Fed_MAPfastR_data$sex.chrom <- NA
Fed_MAPfastR_data$heterogam <- 1
Federico Calboli
On Thursday, 21 March 2013 10:12:39 UTC, Ronnie Nelson wrote: