Identify SNP intervals

[bylemcsmith]

I was wondering what the best tool is to use if you have a set of SNPs and a list of intervals on chromosomes, how do you determine how many SNPs are in each interval? So far I have been writing my own code, which has been very inefficient. It takes days to run. I was wondering if there is a program that will do this?

[rbagnall]

If you have the chromosome co-ordinates of your SNPs, you could use Bedtools.

Save your list of SNPs in bed format as SNP.bed

chr1 100 101 rs1
chr1 105 106 rs2
chr1 110 111 rs3
chr1 5000 5001 rs_not_in_interval
chr2 100 101 rs4
chr2 105 106 rs5
chr2 110 111 rs6
chr2 120 121 rs7
chr2 400 401 rs_not_in_interval


Save your list of intervals in bed format as Intervals.bed

chr1 99 120
chr2 11 130

Then use bedtools intersectBed:

intersectBed -a SNP.bed -b Intervals.bed -wb >SNPs.in.intervals.bed

This will return all SNPs that are in the intervals, and show which interval they are in:

chr1 100 101 rs1 chr1 99 120
chr1 105 106 rs2 chr1 99 120
chr1 110 111 rs3 chr1 99 120
chr2 100 101 rs4 chr2 11 130
chr2 105 106 rs5 chr2 11 130
chr2 110 111 rs6 chr2 11 130
chr2 120 121 rs7 chr2 11 130

To go one further and count how many SNPs are in each interval:

intersectBed -a SNP.bed -b Interval.bed -wb | awk -F"\t" '{print$5" "$6" "$7}' | uniq -c

3 chr1 99 120
4 chr2 11 130

First column gives counts of SNPs in each interval

[crazyhottommy]

that's what I usually do

If you have the chromosome co-ordinates of your SNPs, you could use Bedtools.

Save your list of SNPs in bed format as SNP.bed

chr1 100 101 rs1
chr1 105 106 rs2
chr1 110 111 rs3
chr1 5000 5001 rs_not_in_interval
chr2 100 101 rs4
chr2 105 106 rs5
chr2 110 111 rs6
chr2 120 121 rs7
chr2 400 401 rs_not_in_interval


Save your list of intervals in bed format as Intervals.bed

chr1 99 120
chr2 11 130

Then use bedtools intersectBed:

intersectBed -a SNP.bed -b Intervals.bed -wb >SNPs.in.intervals.bed

This will return all SNPs that are in the intervals, and show which interval they are in:

chr1 100 101 rs1 chr1 99 120
chr1 105 106 rs2 chr1 99 120
chr1 110 111 rs3 chr1 99 120
chr2 100 101 rs4 chr2 11 130
chr2 105 106 rs5 chr2 11 130
chr2 110 111 rs6 chr2 11 130
chr2 120 121 rs7 chr2 11 130

To go one further and count how many SNPs are in each interval:

intersectBed -a SNP.bed -b Interval.bed -wb | awk -F"\t" '{print$5" "$6" "$7}' | uniq -c

3 chr1 99 120
4 chr2 11 130

First column gives counts of SNPs in each interval

[Jeremy]

If you know any R then the GenomicRanges package does that using countOverlaps and or findOverlaps. You can then subset the SNPs based on the results.