To celebrate the release of GATK 4.0, we are giving away free credits for running the GATK4 Best Practices pipelines in FireCloud, our secure online analysis portal. It’s first come first serve, so sign up now to claim your free credits worth \$250. Sponsored by Google Cloud. Learn more at https://software.broadinstitute.org/firecloud/documentation/freecredits

# Haplotype Scoring Algorithm

PanamáMember
edited November 2012

Hi there,

I'm trying to understand the haplotype scoring algorithm in GATK 1.6.5. I fortunately got a printed page where I have a simple diagram that explains the algorithm, I can't find it anymore in the new web.
The case is that the formula for calculating the haplotype score in this diagram has a variable that I'am missing what it is. This is the formula as it's written:

``````P(read | haplotype_j) = sum_bi (bi == hi ? ei : 1 - ei / 3) - sum_bi (ei)
``````

I guess bi stands for base at position i at the current read and hi stands base at position i at haplotype_j, that makes sense for me. But, what is ei?? maybe I'm missing something... it looks like it should be a probability in the range (0, 1) for the haplotype score to make sense.

Pablo.

Post edited by Geraldine_VdAuwera on
Tagged:

• PanamáMember

Hi,

They are not really talking about haplotype scoring algorithm in that article. Anyway that lead me to the fragment-based SNP calling slides and they are referring to an "e" which is the sequencing error rate. May it be this?
We had ei, so it would be specific to that position and not a static sequencing error. I guess it is the error rate for position i in haplotype j, that might be the number of mismatches to consensus haplotype j at position i over the total counts for position i at haplotype j.

But I still doubt how are you calculating the error sequencing rate? Going through the documentation for ErrorRatePerCycle, the error sequencing rates calculated there do not match only to mismatches/counts.

Thanks!
Pablo.

• PanamáMember

Aaaaaaahhhh, OK. I knew it had to be something evident... that makes sense.
I just wanted to fully understand the haplotype scoring and I missed what this e was, I was thinking about calculating some error probability