README.md

In ShamamW/PHCPImpute: PHCP-BASED IMPUTATION

PHCPImpute

An R package for imputation of pseudo-haplotypes data, based on the PHCP algorithm. \ The main function takes a pseudo-haplotype target and a set of reference genomes and returns the posterior probabilities for the possible genotypes in each variant. Full details described in Waldman et al., Genome-wide data from medieval German Jews show that the Ashkenazi founder event pre-dated the 14th century, biorxiv, 2022.05.13.491805 (2002).

Installation

The package was built under R version 4.1.0 \ To install the current packages, the devtools package should be installed. \ Additional required packages for usage: dplyr.

# If devtools is not installed
install.packages("devtools") 
devtools::install_github("ShamamW/PHCPImpute")
library(PHCPImpute)
#if dplyr is not installed
install.packages("dplyr")
library(dplyr)

Usage

phcp_impute(haps, ancient_tped, genetic_map, chr, choose_donors = F, variants)

Input

Required arguments

• haps - a data frame with phased reference genomes in .haps format, which is the output of SHAPEIT2. I.e., the first five columns are: \   1. chromosome number without characters, e.g., 1 and not chr1 \   2. variant ID \   3. physical position in base-pairs \   4. Allele 0 \   5. Allele 1 \ The next columns are the phased haplotypes.

• ancient_tped - a data frame with the target pseudo-haplotype sequence in a Plink tped format. I.e, the first four columns are: \   1. chromosome number without characters \   2. variant ID \   3. genetic distance \   4. physical position in base-pairs \ The next two columns are two identical columns of the pseudo-haplotype sequence.

• genetic_map: Genetic map for the reference genomes in Morgans. The file should have 3 columns: \   1. chromosome name without characters\   2. genetic distance (in Morgans) \   3. position (in bp)

• chr: integer. the number of chromosome to impute, without characters, e.g., 1 and not chr1

• choose_donors: the number of donors' haplotypes to choose from the reference data. If set to FALSE, all donors will be used (default: FALSE).

Optional arguments

• variants: a data frame with a list of variants from the reference panel to impute. If not specifiec, all variants in the refernce panel will be imputed. \ The file should have the following columns with the following headers (additional columns will be ignored): \   1. chr - chromosome number without characters \   2. bp - physical position in base-pairs \   3. Ref - reference allele \   4. Alt - alternate allele
• freq_donors: the maximal minor allele frequency in variants that are used for ranking and choosing the most informative donors. Variants with higher allele frequency will not be used for ranking the donors but will be used in the imputation (default: 0.5).

Output

The output is a data frame with the posterior probabilities of the genotypes in the imputed variants. I.e., each variant is assigned with the probabilities for Ref/Ref, Ref/Alt and Alt/Alt alleles. \ If a data frame with variants to be imputed was supplied to the function ("variants" argument), the output will include only the posterior probabilities of the chosen variants. Otherwise, the output will include the posterior probabilities for all variants in the reference panel.

ShamamW/PHCPImpute documentation built on Oct. 26, 2022, 6:34 a.m.