New version of Python client released (1.0.1)

A new version of the Synapse Python Client was released to PyPi for public consumption. As always, to see the current development of the Python Client, see the source code on GitHub.

Advertisements

FasterCures Webinar On Crowdsourcing and DREAM Challenges

“The way biomedical research is carried out is changing fundamentally,” Sage Bionetworks President Stephen Friend declared at the beginning of a webinar about the crowdsourced computational challenges Sage is facilitating in partnership with the DREAM (Dialogue for Reverse Engineering Assessment and Methods) project that originated at IBM. Friend laid out five opportunities he believes are giving rise to new ways to generate, analyze, and support new research models:

– It’s now possible to generate massive amounts of human “omic’s” data.
– Network modeling approaches for diseases are emerging.
– Information technology infrastructure and cloud computing capacity allow an open approach to biomedical problem solving.
– There’s an emerging movement for patients to control their own sensitive information, allowing sharing.
– Open social media allows citizens and experts to use gaming to solve problems.

“The usual rule of anointed experts being the only ones who can solve problems has really been shattered,” said Friend.

For several years, Sage has been grappling with how to bring about a better understanding of the complexity of biology, given these trends. One initiative central to their efforts has been the creation of a technology platform for data sharing and analysis called Synapse, built on the model of “github” from the open-source software world, which allows distributed projects to get done and provides the foundation for running the DREAM Challenges.

Friend noted that computational biology has been driven by crowdsourcing for a long time, and challenges like those that DREAM has been running for many years have been integral to its successes. There are increasingly large and powerful sets of data in the public domain, and putting them out for many people to look at (some of them from outside the field of biology) and make predictions and unbiased evaluations based on the data is critical to solving complex problems in biology in this day and age. Data is getting so complex that it’s impossible for any single researcher or institution to analyze it effectively. As John Wilbanks, Sage’s Chief Commons Officer and a FasterCures Senior Fellow, noted, “One of the hardest things to do in the emerging Big Data world is to get your data analyzed.”

An important aim of these challenges is to foster a new culture in research. As Friend argues, “We have a serious need not just to solve specific problems, but … to build communities so that people begin to think of each other as colleagues and collaborators.” DREAM Challenges are carefully constructed to provide opportunities for publications in journals and for other forms of recognition that are important to researchers, often more important than the promise of a monetary prize.

First of the four past challenges run by Sage and DREAM (along with partners from academia, industry, government, and patient groups) was the Breast Cancer Prognosis Challenge, created to forge a computational model that accurately predicts breast cancer survival. The winning team was from the academic lab that invented the MP3 format for digital audio, bringing their expertise in data compression to the task. Hundreds of teams comprised of thousands of individuals have participated, and a number of publications have resulted, along with other opportunities for professional advancement for “solvers.”

Challenges currently open include:

– The Somatic Mutation Calling Challenge, to predict cancer-associated mutations from whole-genomic sequencing data;
– The Rheumatoid Arthritis Responder Challenge (in partnership with the Arthritis Foundation, among others), to predict which patients will not respond to anti-TNF therapy – a clinical trial could follow if a powerful classifier emerges from the Challenge for validation; and
– The Alzheimer’s Disease Big Data Challenge, which seeks to predict early AD-related cognitive decline and the mismatch between high amyloid levels and cognitive decline. Massive amounts of data in the public domain has been aggregated, collated, massaged and curated for the task.

Two more are set to open this summer, in partnership with the Broad Institute and MD Anderson Cancer Center, and several more are being considered for launch by the end of 2014. All stakeholders – including and perhaps especially patient groups – are invited to participate by proposing ideas for challenges, contributing data, recruiting teams to participate. The Sage-DREAM Challenges are looking for partners who want not only to find the answers to tough questions in their fields, but who want to help create the conditions for the real collaboration necessary to bring about “the next generation of biomedical research.”

For more information on how to get involved with an open DREAM Challenge, click here.

View webinar slides and recording

(Cross posted from http://fastercures.tumblr.com/post/81603549119/crowdsourcing-data-challenges-to-speed-the-search-for)