GATK- haplotypecaller or unifiedgenotyper?

[lre1234]

Hi all,
I have been looking at the GATK documentation at in the best practices it says:

We believe the best possible caller in the GATK is the HaplotypeCaller

I don't really understand why the haplotype caller is better than the unifiedgenotyper. If I am going to be doing 2 whole genomes (~40X each) on a paired tumor/normal, is the haplotypecaller the best option? I had used the unifiedgenotyper in the past.

Thanks

[sdavis]

To answer the question you are really asking and not the question in the title, if you are looking for somatic mutations, then neither is likely to be the best tool. You might consider looking at somatic variant callers such as muTect, strelka, somaticsniper, or VarScan2.

[BurlEarl]

The Tumor is not going to be diploid. This will be incompatible with HaplotypeCaller and high ploidy slows UnifiedGenotyper way way down. Just make sure your caller is meant for pooled or high pliody samples. As for the above suggestions, i haven't used them. I use SNVer right now for viral population projects, it seems to go the trick and is painless to get up and running.