Do Not Blink

2014-05-20

Blink and you will miss a revolution! This is indeed true in the field of biomedical informatics today as it transforms healthcare and clinical translational research at light speed. Two recent conferences – AMIA’s Translational Bioinformatics (#TBICRI14) and BioIT World (#BioIT14) brought together national and international informatics experts from academia, industry and non-profit organizations to capture a snapshot of scientific trends, marvel at the progress made and the opportunities ahead. I hope to give you a glimpse of my personal journey at these two conferences with references to additional information should you decide to delve deeper.

Stanford Professor and creator of PharmGKB, Russ Altman’s (@Rbaltman) presented the year in review, a cherished event at the annual AMIA TBI conference, which highlighted the 42 top papers in the field. He started with the warning letter from the FDA to Ann Wojcicki, CEO, 23andMe to stop marketing the consumer gene testing kit that is not FDA cleared; and followed with a Nature commentary by Robert Green and Nita Farahany asserting that the FDA is overcautious on consumer genomics. The authors cited data from over 5000 participants that suggested that consumer genomics does not provoke distress or inappropriate treatment. Russ then reviewed Harvard professor John Quackenbush’s (@johnquackenbush) Nature Analysis paper, which showed that the measured drug response data from two large-scale NIH funded pharmacogenomics studies were highly discordant with large implications for using these outcome measures to assess gene-drug interactions for drug development purposes. Large-scale database curation related shout outs included Pfizer-CTD’s manual curation of 88,000 scientific articles text mined for drug-disease and drug-phenotype interactions published by David et al., in the Journal Database; and the DGIdb: mining the druggable genome by Griffith et al., in Nature Methods. Russ’s crystal ball for 2014 predicts an emphasis on non-European descent populations for discovery of disease associations, crowd-based discovery using big data, and methods to recommend treatment for cancer based on genomes and transcriptomes.

The 7 AM birds-of-a-feather session on “researching in big data” facilitated by Columbia’s Nick Tatonetti (@nicktatonetti) engaged a vocal group of big data proponents where we discussed the definition (the four V’s – velocity, volume, veracity and variety), processing (MapReduce/Hadoop, Google APIs), visualizing (d3.js) and sharing of massive biomedical datasets (no-sql databases to cloud based resources). New analytical tools were presented including netGestalt from Vanderbilt for proteogenomic characterization of cancer; PhenX toolkit from NIH for promoting data sharing and translational research; SPIRIT from City of Hope for protocol decision trees, eligibility screening, and cohort discovery and many others.

Method presentations included an integrated framework for pharmacogenomics characterization of oncological drugs, and novel NGS analysis methods on the Amazon cloud, by ICBI members Krithika Bhuvaneshwar and Brian Conkright, respectively. A keynote lecture given by Richard Platt described PCORI’s PCORnet coordinating center, a newly established consortium of 29 networks that will use electronic health data to conduct comparative effectiveness research and the 18-month schedule to get the consortium up and running. Zak Kohane, Director, Center for Biomedical Informatics at Harvard Medical School kicked his keynote lecture out of the park again. He described, among other things, the critical role of translational bioinformaticians in translating big data to clinically usable knowledge. The entire conference proceedings can be found here.

Last week, BioIT World started with an excellent keynote by John Quackenbush where he described his journey as the co-founder of Genospace. As digital architects of genomic medicine, the company aims to improve the progress and efficacy of healthcare in the genomic age. John finished his talk by emphasizing that the most important “omics’ in precision medicine is econOMICS, especially given that the first slide of most everyone who talks about precision medicine these days is one that shows the drop in cost per megabase of DNA sequence compared to Moore’s law. On the other hand, Stephen Friend, President of Sage Bionetworks, discussed during his keynote provocative questions such as “why not have a GitHub for data?” and “can we have sponsors such as Gates and NIH push open data access for programs they fund?” BioIT world this year had 13 parallel tracks covering a wide range of topics from cloud computing to cancer informatics.

I attended talks including transSMART – a community driven open source platform for translational research by Roche and the transSMART foundation; the Pan-Cancer analysis of whole genome projects by Lincoln Stein of the Ontario Institute for Cancer Research; and MetaboLYNC – a cloud based solution for sharing, visualizing, and analyzing metabolomics data by the company Metabolon, Inc. Diagnosis and treatment in elderly patients presents a unique set of challenges because of their extensive clinical history, altered physiology and physiological response both to diseases and treatments, patterns of behavior and access to appropriate medical care. A talk by Michael Liebman, IPQ analytics and Sabrina Molinaro, Institute of clinical physiology, Italy highlighted the application of big data to address the complexity in treatment of elderly patients with diabetes and hypertension.

I had the wonderful opportunity to chair a session on “Clinical genomics data within cloud computing environments” and shared our experience at Georgetown as we build a cancer cloud in collaboration with University of Chicago and the Globus Genomics team.

With so many exciting talks and demonstrations of terrific progress in informatics at both conferences, I did feel that I could not blink lest I should miss something of extreme importance. I welcome you to check out the rest of the newsletter to catch up on exciting events and activities at ICBI. Let’s continue the conversation – find me on e-mail at sm696@georgetown.edu or on Twitter at @subhamadhavan.