Identify the nearest common SNP to a given rare variant

Question

[bisansamara]

I have a list of rare single nucleotide variations (SNVs), and I would like to identify the nearest common SNP to each from both ends.

---

The input information would be the coordinates of the SNVs (something like this: chr7:127991052, chr18:321720, chr5:76174154, etc.). The output I need is the nearest SNP to each SNV (both upstream and downstream).

---

What tools could perform this task? Your help is much appreciated!

Answer 1

[Pierre Lindenbaum]

assuming the file is sorted

grep -A 1 -B 1 -F 'chr7:127991052' input.txt

Answer 2

[Alex Reynolds]

Via BEDOPS:

$ vcf2bed < list.vcf > list.bed
$ vcf2bed < snps.vcf > snps.bed
$ closest-features list.bed snps.bed > answer.bed

[bisansamara]

This seems like a good tool. But it's taking forever to convert the dbSNP file (downloaded from NCBI) from csv to bed....any suggestions? Note that I'm using Cygwin on Windows to run the commands.

[Alex Reynolds]

To get around limitations in /tmp you could specify an alternative temporary directory for sorting. Add --sort-tmpdir=<dir> and (for example) --max-mem=2g to do sorting in a non-temporary (non /tmp) directory and to sort with 2GB of system memory. See --help or the online documentation for more detail.