How to get TCGA SNP in vcf and process in PLINK

Question

[Omics data mining]

Dear all

I urgently want to download SNP information of TCGA-PAAD and run in PLINK for further analysis . I got the expression data of 183 samples. Is It possible to get SNPs of 183 samples in VCF ?

waiting for reply

All suggestions will be appreciated

Thank you in advance

Archana

Answer 1

[pbpanigrahi]

Download the snp information as maf format, then you can convert maf to vcf format using https://github.com/mskcc/vcf2maf/blob/master/maf2vcf.pl

[Omics data mining]

Hi

I did the same. During runtime of maf2vcf.pl I am getting warning

WARNING: Reference allele mismatches found.

Many of SNPs are getting filtered out . What should I do ?

[Omics data mining]

I checked respective position on given genome build, It was exactly same. Still I am getting this issue. Can anybody suggest me how to fix it ? Otherwise I will lose most of the SNPs information for downstream which I do not want to do. Or If its possible to create manual vcf ? How ?

[pbpanigrahi]

Can u post the command u typed. Also are u getting many vcf files in the output folders? Because the code generates a vcf file per tumor-normal sample pair. Are u getting only the specific error u mentioned earlier, or other error. Detailed information will help us to debug. Else provide the input file.

[Omics data mining]

Hi

First of all, defined the ref fasta my $ref_fasta = "/path/Homo_sapiens.GRCh37.75.dna.toplevel.fa" in script maf2vcf.pl and saved changes. Here is command used to convert maf to vcf .

perl maf2vcf.pl --input-maf TCGA-IB-8127-01.maf.txt --output-dir TCGA-IB-8127-01.vcf

Thank you in advance Archana