snpTools

My R package for general tasks when working with SNP array data

Table of Contents

1. Summary of tools
2. Installation
3. Example

Summary of tools

Currently, snpTools contains:

• merge_geno() Merges any number of genotyping datasets and ensures SNPs from each set are merged in the correct order.
• recode_geno() Converts an ACTG coded SNP to a A/B coded SNP, provided a recoding table. The function can be vectorized to apply to a full dataset of n SNPs
• count_geno() Converts a SNP dataset in A/B format to the dosage of a chosen allele.
• filter_geno() Filters a SNP dataset, filtering out SNPs that do not meet a particular call rate, are fixed, or have an extreme MAF (to be implemented).
• format_gpGeno() Imports raw genotyping data (in the format such as those provided by GeneSeek)
• fimpute_run() Runs FImpute from within R, allowing for convenient genotype phasing and imputing from genotyping data and a SNP map.

Installation

To install and attach snpTools, from within the R interpreter use:

devtools::install_github("funkhou9/snpTools")
library(snpTools)

This assumes you have the devtools package installed. If not, use install.packages("devtools") to install devtools from CRAN.

For more information on any particular function, use:

?fimpute_run

Example

fimpute_run() can be used to phase genotypes, given a genotyping dataset and SNP map with:

fimpute_run(geno = <geno>,
            map = <map>,
            path = "/path/to/FImpute",
            exclude_chr = "1 2 3 4 5 6 7 9 10 11 12 13 14 15 16 17 18 19 20 21")

• <geno> may be a data.frame or matrix containing SNPs in columns and individuals in rows. SNPs must be represented as dosages {0, 1, 2}. Rownames must contain the name of each individual.
• <map> must be a data.frame containing SNP information, with columns labeled as "chr" and "pos", and rownames containing the name of each SNP.
• Use the path argument to specify the location of the FImpute binary.
• Use the exclude_chr to specify a character string containing the chromosomes you want to exclude in the genotype phasing. In the example we are only phasing SNPs present on chromosome 8.
• Use the optional output_folder argument to specify the name and location of FImpute output. By default, output will be saved in the current working directory within a directory named fimpute_run

The above code will invoke FImpute, so you should see standard FImpute stdout:

 .--------------------------------------.
|  .---.--.--                  .         |
|  |      |                   _|_        |
|  |---   |  .--.--. .,-. .  . |  .-.    |
|  |      |  |  |  | |   )|  | | (.-'    |
|  '    --'--'  '  `-|`-' `--`-`-'`--'   |
|                    |                   |
|                    '                   |
|              Version 2.2               |
|                                        |
|           Mehdi Sargolzaei,            |
|  Jacques Chesnais and Flavio Schenkel  |
|           Semex Alliance, ON           |
|     CGIL, University of Guelph, ON     |
|                                        |
|       Last update: Jan 16, 2014        |
 `--------------------------------------'