November 11, 2013
Janet Janjigian, Carmen Group West
Marina del Rey, California- November 11, 2013 – Heritage Provider Network, (HPN) The National Cancer Institute, (NCI) Sage Bionetworks and DREAM announced the best performers in the Heritage-DREAM Breast Cancer Network Inference Challenge. The competition challenged teams to use big data to develop more effective treatments for breast cancer. Six best performing teams were honored with cash prizes at the Annual RECOMB/ISCB Conference on Regulatory and Systems Genomics, with DREAM Challenges in Toronto, Ontario on November 8, 2013. Sponsored by HPN and the NCI Division of Cancer Biology, (DCB) the goal of the competition was to challenge data scientists to find ways to increase our understanding of signaling cellular pathways in breast cancer cells that could lead to more effective treatments for breast cancer patients.
Six teams were chosen as best performers within three categories of the challenge. Category questions posed to participants included (1) identifying cellular pathways that transmit extra cellular signals in breast cancer cells (2) predicting the dynamics of the proteins responsible for transmitting extracellular signals in breast cancer cells and (3) how to visualize complex big data sets in biomedical research such as those used in this challenge.
“I am so pleased with this successful collaboration between HPN and the NCI in sponsoring this important competition,” said Dr. Richard Merkin, President and CEO of HPN. “More than two hundred teams completed to better predict the networks and signaling dynamics in breast cancer cells, with one hundred teams making the final submission round. The innovations fostered by this competition will undoubtedly lead to an improved understanding of breast cancer, identify new therapeutic targets and most importantly, will help save lives. Congratulations to the six best performing teams.”
“This was a complex and difficult challenge, dealing with the way breast cancer cells process extracellular information to alter their internal states,” said Gustavo Stolovitzky, co-founder of the Dream Project. “When we formulated the challenge, there was no certainty that there would be any teams that could solve the problem to any degree of significance. We are extremely pleased that several teams managed to find methodologies that yielded pretty accurate solutions.”
“Cancer remains one of the most challenging and devastating diseases afflicting the planet,” said Dan Gallahan, Deputy Director, DCB, National Cancer Institute. “The NCI has worked tirelessly to develop new understandings of cancer with the direct goal of reducing its burden. Besides exploring new scientific discoveries, the NCI is also exploring new methods to uncover those discoveries. Crowd sourcing and challenges represent a new opportunity to go beyond our traditional research paradigm and engage a whole new community in a new way,” he continued. “The DREAM program has been at the forefront of developing these types of research activities for the biomedical community and likewise, HPN has been a generous leader in the support of biomedical research. We would like to thank both these groups for making this important challenge possible and advancing our knowledge of breast cancer. I would also like to personally congratulate the top performers and all participants in the Heritage- DREAM challenge.”
Best performers include Team DC-TDC from the University of California at Santa Cruz, Team NMSUSongLab from New Mexico State University, Team GuanLab from the University of Michigan, Team StochasticChaos from Johns Hopkins University Sidney Kimmel Comprehensive Cancer Center and the University of North Carolina at Chapel Hill, Team CGR from NCI and the Chinese Academy of Medical Sciences in Beijing and Team ABCD from Rice University.
“Team DC-TDC is delighted to have contributed a winning solution by identifying cellular pathways that transmit extra cellular signals in breast cancer cells,” said Josh Stuart, team member. “The DREAM competition is an incredible motivating force and a truly fun experience for all of us. In the end, we came up with a very clever strategy that combines the use of heat diffusion on a “SuperPathway” of known genetic knowledge and a twist on a theory borrowed from economics called Granger Causality. The reward for us is that it pushed us to achieve a new level of creativity in our approaches to reveal cancer processes. As always, we remain hopeful that our efforts as a community of friendly competitors will lead to contributions in fighting this disease.”
The competition will now move into a collaborative phase where teams will be able to share information and work together to find solutions. The deadline for this next phase of the challenge is January 15, 2014.