Open Genomics is coming together
Here's a quick history of our community, its motivation, the people who started it, and plans for the future at the moment of inception.
It all started with John St John of Driver Genomics connecting around Scala By the Bay 2014, and sharing his plans about using Scala, Play, and Spark to organize genomics data. I've learned a lot, shared some pointers, and we welcomed several Driver developers at the conference, and its Fast Track to Scala and Fast Track to Spark courses -- the first run together with Typesafe, the latter in together with Databricks. Stewart Stewart from Driver (yes, Stewart^2) took the Spark one.
Come 2015, Stewart Stewart organized an SF Scala Puzzle Night at Driver's office near UCSF Mission Bay. As an intro to Driver, we got a tour of the facilities. A part of it was the cofounders telling us what they do and why they do it. A lot of people got quiet listening to it, seeing cancer patients on screen who could have been treated if personalized medicine existed -- but whose genomes weren't even sequenced. I've been personally moved by those stories, and realized that Driver's work, with broader set of related efforts in biotech, including AMPLab, is the best way forward to defeat cancer faster. Having started and run four open-source meetups -- SF Scala, SF Spark, SF Text, and Reactive Systems -- I know how to mobilize developer communities, and I've immediately resolved to help Open Genomics tribe to come together from our meetups and broader Bay Area.
John recommended Matt Massie's group at AMPLab as the next stop. I've visited Matt with his colleague Frank Nothaft, and once we learned that we all prefer Spark for compute and Parquet+Avro as our data format, we'd realized we were brothers. We planned a series of joint activities, including talks at the coming Text By the Bay conference, and a working group meeting preceding it.
It turns out Natural Language Processing, which is all about strings, developed technologies directly applicable to genomics. UPenn, one of the leaders in NLP, has a long history of knowledge transfer between Computational Linguistcs and Bioinformatics. My UPenn Ph.D. thesis advisor, Prof. Lyle Ungar, and a member of my Ph.D. committee, also the Director of Linguistic Data Consortium, Mark Liberman, were both speakers at Text By the Bay, with Mark, one of the key authorities in the field, delivering the opening keynote. We all met on April 23, 2015, at Galvanize, for a whole-day working group formation. Other members included John St John of Driver, Frank Nothaft of AMPLab, Jeff Lerman of QIAGEN, Malcolm Greaves from Nitro, Mike Tamir and Nir Kaldero from Galvanize.
We discussed what's needed to help OSS developers help fight cancer. They need to ramp up on biology first, which is more than perusing a README file. Conversely, for biologists to get to the point of appreciating why Parquet+Avro is a reasonable data format, probably takes a similar ramp. We decided to continue working together, and hold a series of meetups to extend it to a group of volunteers who will help with gathering the resources needed.
Frank's and John's talks at Text By the Bay were fantastic, and had the same effect on the conference attendees as they had on me.
The first Open Genomics meetup was held at SF Spark and Friends meetup on July 23rd, 2015, at Chartboost, and our signup form had 20 members joining. The path forward will include a stable group of volunteers gathering links and resources on the web site, collaboration with universities and startups on hackathons, a Kaggle competition -- or a series -- and perhaps forming a nonprofit around this work.
As a meetup organizer, I want to dedicate my time and effort to Open Genomics community as a way of giving back, and our working group is looking for a standing committee of volunteers to both move this forward and decide the formal shape of this work. Please sign up for the group and we'll make progress together.
kaledo-art
izzy806
dailybridgerton