comparing vcf files of disease and normal samples

Question

[GouthamAtla]

Hi ,

I have two VCF files that belongs to disease samples and four VCF files that generated from normal samples. Now I would to filter of SNPs . I would like to get the SNPs that are present in only disease group so that I can go further to validate them by other methods. Is there any tool that does this kind of analysis?

Answer 1

[Katie D'Aco]

I would start with vcftools merge command and then write a script to remove variants that are present in controls. Maybe there is an existing tool that will filter by genotype call in certain samples, but I am not aware of one.

If the disease you're looking at is rare, I would also consider filtering by 1000 genomes or NHLBI allele frequencies.

[GouthamAtla]

Thank you for the reply. I have used multisample SNP calling ( UnifiedGenotyper ) and filtered SNPs and selected SNPs with AWK.

[QPaps04]

hello geek_y - I realise this is a very old post but please can you explain how you used AWK to select SNPs?

Answer 2

[Vivek]

You can start with GATK CombineVariants to merge all the VCFs at once.

Then use GATK SelectVariants to filter out variants that occur in only the two disease samples.

[GouthamAtla]

Thank you for the reply. I have used multisample SNP calling ( UnifiedGenotyper ) and filtered SNPs and selected SNPs with AWK.

Answer 3

[ramanandan.prabhakaran]

Hi Goutham,

My requirement is also some what similar to yours. I have to find snps which are present in affected but not in unaffected samples, then I have to prepare a list For example,

Samples           4:1243-SNV     5:1277-SNV       15:4070-SNV    ..... ..... ...... ...... ..... .....      16:5335-SNV 
A(affected)          C_T                  A_G                 A_T               ..... ..... ...... ...... ..... .....        A_C
B(unaffected)      C_C                  A_A                  A_A               ..... ..... ...... ...... ..... .....       A_A
C(affected)          C_T                  A_G                 A_T               ..... ..... ...... ...... ..... .....        A_C
D(affected)          C_T                  A_G                 A_T              ..... ..... ...... ...... ..... .....         A_C

Any help :)

Cheers,
Ram