The current GATK version is 3.7-0
Examples: Monday, today, last week, Mar 26, 3/26/04

#### Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

#### ☞ Formatting tip!

Surround blocks of code, error messages and BAM/VCF snippets--especially content with hashes (#)--with lines with three backticks ( ` ) each to make a code block.
GATK 3.7 is here! Be sure to read the Version Highlights and optionally the full Release Notes.

# GATK 3.0 Features Overview

edited February 2014

Previously, we covered the spirit of GATK 3.0 (what our intentions are for this new release, and what we’re hoping to achieve). Let’s now have a look at the top three features you can look forward to in 3.0, in no particular order:

1. Optimized PairHMM algorithm to make GATK run faster
2. Single-sample pipeline for joint variant discovery
3. Best practices for calling variants on RNAseq data

### 1. Optimized PairHMM algorithm to make HaplotypeCaller faster

At this point everyone knows that the HaplotypeCaller is fabulous (you know this, right?) but beyond a certain number of samples that you’re trying to call jointly, it just grinds to a crawl, and any further movement is on the scale of continental drift. Obviously this is a major obstacle if you’re trying to do any kind of work at scale beyond a handful of samples, and that’s why it hasn’t been used in recent large-cohort projects despite showing best-in-class performance in terms of discovery power.

The major culprit in this case is the PairHMM algorithm, which takes up the lion’s share of HC runtime. With the help of external collaborators (to be credited in a follow-up post) we rewrote the code of the PairHMM to make it orders of magnitude faster, especially on specialized hardware like GPU and FPGA chips (but you’ll still see a speedup on “regular” hardware).

We plan to follow up on this by doing similar optimizations on the other “slowpoke” algorithms that are responsible for long runtimes in GATK tools.

### 2. Single-sample pipeline for joint variant discovery

Some problems in variant calling can’t be solved by Daft Punk hardware upgrades (better faster stronger) alone. Beyond the question of speed, a major issue with multi-sample variant discovery is that you have to wait until all the samples are available to call variants on them. Then, if later you want to add some more samples to your cohort, you have to re-call all of them together, old and new. This, also known as the “N+1 problem”, is a huge pain in the anatomy.

The underlying idea of the “single-sample pipeline for joint variant discovery” is to decouple the two steps in the variant calling process: identifying evidence of variation, and interpreting the evidence. Only the second step needs to be done jointly on all samples, while the first step can be done just as well (and a heck of a lot faster) on one sample at a time.

The new pipeline allows us to process each sample as it comes off the sequencing machine, up to the first step of variant calling. Cumulatively, this will produce a database of per-sample, per-site allele frequencies. Then it’s just a matter of running a joint analysis on the database, which can be done incrementally each time a new sample is added or at certain intervals or timepoints, depending on the research needs, because this step runs quickly and cheaply.

We’ll go into the details of exactly how this works in a follow-up post. For now, the take-home message is that it’s a “single-sample pipeline” because you do the heavy-lifting per-sample (and just once, ever), but you are empowered to perform “joint discovery” because you interpret the evidence from each sample in light of what you see in all the other samples, and you can do this at any point in the project timeline.

### 3. Best practices for calling variants on RNAseq

Our Best Practices recommendations for calling variants on DNA sequence data have proved to be wildly popular with the scientific community, presumably because it takes a lot of the guesswork out of running GATK, and provides a large degree of reproducibility.

Now, we’re excited to introduce our Best Practices recommendations for calling variants on RNAseq data. These recommendations are based on our classic DNA-focused Best Practices, with some key differences the early data processing steps, as well as in the calling step. We do not yet have RNAseq-specific recommendations for variant filtering/recalibration, but will be developing those in the coming weeks.

We’ll go into the details of the RNAseq Best Practices in a follow-up post, but in a nutshell, these are the key differences: use STAR for alignment, add an exon splitting and cleanup step, and tell the variant caller to take the splits into account. The latter involves some new code added to the variant callers; it is available to both HaplotypeCaller and UnifiedGenotyper, but UG is currently missing a whole lot of indels, so we do recommend using only HC in the immediate future.

Keep in mind that our DNA-focused Best Practices were developed over several years of thorough experimentation, and are continuously updated as new observations come to light and the analysis methods improve. We have only been working with RNAseq for a few months, so there are many aspects that we still need to examine in more detail before we can be fully confident that we are doing the best possible thing. We will be improving these recommendations progressively as we go, and we hope that the researcher community will help us by providing feedback of their experiences applying our recommendations to their data.

Geraldine Van der Auwera, PhD

Tagged:

• Cambridge, UKMember Posts: 111 ✭✭✭

Am I correct in my reading of your poster that you've rewritten a fair ol' bit of haplotype caller in c++?

Martin Pollard, Human Genetics Informatics - Wellcome Trust Sanger Institute and Genetic Epidemiology Group - WTSI & Cambridge University

You'll have to ask @Carneiro for the gory details (or wait for the upcoming in-detail post), but as far as I know it took quite a bit of re-implementing things in c++, yes. Though strictly speaking it's the PairHMM code that was rewritten (which is also used by UG), not the HaplotypeCaller overall.

Geraldine Van der Auwera, PhD

• ConnecticutMember Posts: 9

Is there any special option that I need to specify in order to take advantage of the GPU optimization?

Apologies but the GPU stuff didn't make it into the 3.0 release. It's this close to being ready though so we are contemplating making a 3.1 release really soon with just this stuff added. Will let you know more when we figure out our plans.

Eric Banks, PhD -- Director, Data Sciences and Data Engineering, Broad Institute of Harvard and MIT

• Member Posts: 29

"GPU stuff ... close to being ready"

Does GPU stuff also include the AVX stuff on Mauricio's poster? If I understand AVX correctly, recent versions of Intel Xeon chips coupled with a recent version of Linux will see a performance boost, without any user intervention.