GenomeRunner uses a background, or universe, of genomic regions for random sampling. For SNPs, the background may consist of all currently reported SNPs (useful for the analysis of SNP sets from Genome-Wide Association Studies), or contain a set of all SNPs on a microarray chip (e.g., ImmunoChip, MetaboChip). Think about the background as all SNPs tested in a GWA study, and the SNP sets (FOIs) as subsets of the background significantly associated with a disease/phenotype.
To create a background set of SNPs, obtain genomic coordinates of all currently reported SNPs found, e.g., in >=1% of samples from the UCSC table browser. For Homo Sapiens they may be obtained by selecting Mammals/Human/hg19/Variation/All SNPs(138)/snp138Common in the appropriate drop-down menus, selecting "BED - browser extensible data" as output format, specifying snp138Common.bed as the file name, and getting output as a gzip compressed file. Place the snp138Common.bed.gz file into the [dir]/custom_data/backgrounds/hg19/ folder.
Other background files, such as (for Homo Sapiens) snp138, snp138Flagged, can also be downloaded and placed in the [dir]/custom_data/backgrounds/hg19/ folder using similar strategy. To prioritize their order in the drop-down menu on the web page, each file name should be prefixed with a number that will be used to sort them, e.g., 1snp138Common.bed.gz.
As with the regulatory datasets, we should ensure that no non-standard chromosome names are present in the background files, and the end coordinates are larger than the start coordinates. We also compress and index the background files.
for file in `find [dir]/custom_data/backgrounds/[org] -type f -name "*.bed.gz"`; do f=`basename $file`; d=`dirname $file`; echo $file; zcat $file | grep "\bchr[0-9XYM][^_]\b" | awk 'BEGIN {OFS="\t"} { if ( $3 <= $2) { print $1, $2, $2+1 } else { print $1, $2, $3 } }' | sort -k1,1 -k2,2n -k3,3n | uniq > $d/${f%???} && rm $file; bgzip ${file%???} && tabix $file; done
Note that we are creating BED3 files, that is, keeping chrom, chromStart and chromEnd columns. The rationale is to avoid duplicates, as some SNPs have the same genomic coordinates but different rsIDs.
The resulting folder with background sets may look like:
/home/testuser/Documents/db_1.01_05.12.2014/custom_data/backgrounds/hg19/1snp138.bed.gz
/home/testuser/Documents/db_1.01_05.12.2014/custom_data/backgrounds/hg19/1snp138.bed.gz.tbi
/home/testuser/Documents/db_1.01_05.12.2014/custom_data/backgrounds/hg19/2snp138Common.bed.gz
/home/testuser/Documents/db_1.01_05.12.2014/custom_data/backgrounds/hg19/2snp138Common.bed.gz.tbi
/home/testuser/Documents/db_1.01_05.12.2014/custom_data/backgrounds/hg19/3snp138Flagged.bed.gz
/home/testuser/Documents/db_1.01_05.12.2014/custom_data/backgrounds/hg19/3snp138Flagged.bed.gz.tbi