You've got basic experience with GATK, MuTect, Picard or related genome analysis tools, and a burning desire to help enable great science for a living? Join us now by applying for job #1774 on the Broad careers web page.
If you're a regular here, you may know I joke about being a glorified tech support monkey sometimes, but let me assure you that this job -- this team -- is about much more than that.
We are currently a team of three -- Sheila (associate computational biologist), David (senior technical writer) and me, Geraldine (bioinformatics scientist, formerly wetlab microbiologist).
As the support and outreach crew for GATK, it's up to us to make sure that the user community -- tens of thousands of researchers worldwide -- has the tools and understanding they need to go out and perform great research. Quite a bit of that research has direct impact on patient outcomes, real human lives (see the user stories for some examples). Some of it follows a more fundamental-knowledge path, and that's great too; it all goes to building up the edifice of scientific knowledge. I sincerely believe that we actively contribute to that research by equipping the community with the right information in the right format.
That is heck of a challenging task. It involves fielding questions from individuals from all sorts of backgrounds (from pure biology to pure computer science) with all sorts of questions (from experimental design to details of algorithms, with a splash of hardware optimization on the side) coming from all sorts of language backgrounds that can complicate communication (I'm originally francophone, myself). And that's just one side of the equation. On the other side, we're working closely with the development team, who are also very heterogeneous: we have software engineers, mathematicians, at least one former hedge fund quant, computational biologists, statisticians, and plenty more besides. All with their own particular perspectives that inform how they communicate about their work, which we then need to boil down into actionable information tailored for the researchers who use the software. So we do a lot of translating of all kinds!
And it's not just back and forth questions and answers or writing documentation, either. Personally, some of the parts that I like best about the job are the opportunities to directly influence the software design and feature set, which typically lead to improved usability and applicability of the tools, based on feedback from users and in-depth discussions with the developers. I always wanted to get into software development and never had any properly relevant qualifications, so it's a real treat to feel involved in that part of the process. Sometimes I even get to contribute minor features or bug fixes myself, which is way cool.
Finally, I can't not mention workshops -- especially the invited workshops held in exotic foreign lands, like, um, Belgium (hey don't knock it, that's where I'm from). Also, Thailand, the UK, and soon, South Africa! These workshops typically involve a team of three people (typically one from the support team and two developers) traveling to the host institution and teaching a multi-day workshop. Some of those who participate originally do it for the free travel; all end up really enjoying the immensely rewarding experience of meeting our users in person. GATK users are on average an absolutely lovely bunch of people, and it's really fun to get to spend a few days interacting, teaching and collecting feedback.
Okay, there's still plenty more to it but I've already gone on too long, so if you want to know more, just apply or message me!
How to apply: got to the Broad careers web page, just enter 1774 in the "requisition number" box and don't worry about the rest.
Here's some good news for anyone who has been using both GATK and the Picard tools in their work -- which means all of you, since you all follow our Best Practices to a tee, right?
As you may know, both toolkits are developed here at the Broad Institute, and are deployed together in the Broad's analysis pipelines. The fact that they have been developed, released and supported separately so far is more an accident of history and internal organization than anything else (and we know it's inconvenient to y'all).
The good news is that we're taking steps to consolidate these efforts, which we believe will benefit everyone. In that spirit, we have been working closely with the Picard tools development team, and we're now ready to take the first step of consolidating support for the tools. From now on, you will be able to ask us questions about the Picard tools, and report bugs, in the GATK forum. And developers will be happy to hear that we are also committed to supporting HTSJDK for developers through the Github repo’s Issues tracker.
In the near future, we will also start hosting downloads and documentation for the Picard tools on the GATK website. And before you ask, yes, the Picard tools will continue to be completely open-source and available to all free of charge.
To recap, we have brought the GATK and Picard teams together, and we are working on bringing together in the same place all the methods and tools to perform genome analysis. Our goal is to make a world where you can run our complete Best Practices pipeline end-to-end with a single Broad toolkit. We think it’ll make your life easier, because it sure is making ours easier.
I'm very happy to introduce Sheila Chandran, our newest GSA team member, who will be helping me with GATK outreach, support and documentation. You can expect to see Sheila start answering questions on the GATK forum within a week or two!
Thanks to Sheila's help, I'll be able to expand our support model to the Broad Cancer Tools (MuTect and related). Moving forward, we'll produce documentation for MuTect and other tools produced by the Cancer Group in collaboration with the developers, in order to bring you the same level of documentation coverage and support that we currently have for GATK.
Rest assured however that we won't stop working on improving the GATK documentation. In fact, Sheila's first project with us will be to document how the HaplotypeCaller works in detail -- something I know many of you have been hoping to see for a while now!
As a final note, I'd like to mention that this is one of many positive outcomes from our collaboration with our commercial licensing partner, Appistry, so we'd like to express our thanks to them for helping us help you, our user community.
We're planning to organize online live Q&A sessions in which you will be able to submit questions (within a predefined topic, eg. "base recalibration" or "variant calling") for the developers to answer. To help us choose between several software options, we would like to know which of the following options would be the most valuable to you (please choose one):
Ability to vote for questions, so that the most popular questions get answered (we expect a lot of questions and will probably not have time to answer all of them).
Ability to chat directly with us if your question is picked, so that if our answer doesn't really answer your question you can clarify or ask for more details.
Feel free to tell us in the comments what other features would be useful.