Human GRCh38 and dbSNP VCF and GATK

Question

[win]

Hi all, I used the GRCH38 reference file and aligned my FASTQ files to this, there were no issues here.

Then i used GATKs BaseRecalibration tool and used dbSNP144, at this step I get an error that states VCF and reference have incompatible contigs.

Could someone point where the issue might be or if there a GATK version of GRCh38 and dbSNP VCF.

Thanks

Answer 1

[Pierre Lindenbaum]

see http://gatkforums.broadinstitute.org/gatk/discussion/3843/plans-to-update-the-gatk-bundle

ftp://[e-mail]/bundle/hg38/hg38bundle/

(march 10) As we've mentioned previously, we are providing a beta version of the hg38 bundle but it is not yet officially supported, it is not guaranteed to be complete, and we are not taking requests for additional files. When the full official bundle is ready, we will make an announcement on the blog and its contents will be documented.

Answer 2

[Carlos Borroto]

You problem seems to be you are using a dbSNP VCF with GRCh37 coordinates. The FTP location Pierre links in his answer has a GRCh38 version that should solve your problem.

However, you should know the current GATK version is not "alt" aware yet. While mappers like BWA are and can correctly map reads against a reference with alt contigs, current GATK will most likely fail to call variants in regions with alt contigs. If you must have GRCh38 and GATK at this time, you should strip the reference of any alt contigs before mapping. I believe this invalidates the main advantage of using GRCh38 and is the reason I'm holding the move to GRCh38.