Sept. 13, 2007 | How good is your algorithm? The DREAM project — Dialogue on Reverse Engineering Assessment and Methods — is giving researchers a chance to find out while DREAM organizers pursue their broader goal of promoting development of more efficient, validated computational tools for assessing biological systems and data.
In late July, DREAM organizers posted five challenges with freely downloadable datasets, and invited researchers to use computational methods to solve them. The challenges include: BCL6-Targets Challenge, the Protein-Protein Subnetwork Challenge, the Five-Gene-Network Challenge, the In-Silico-Network Challenge, and the Genome-Scale Network Challenge.
DREAM is loosely modeled after CASP (Critical Assessment of Techniques for Protein Structure Prediction), the biennial competition to computationally predict the precise 3-D shape of an array of target proteins from their amino acid sequences. Now, DREAM is gearing up for its second meeting (DREAM2) planned for December 3-4 at the New York Academy of Sciences.
Results from the first set of challenges will be presented at DREAM2, and short papers describing the best prediction will be included in the Proceedings of the Second DREAM Conference. Accepted papers accompanying the best predictions will also be considered for publication in the journal Molecular Systems Biology.
DREAM2’s “objective is to catalyze the interaction between experiment and theory in the area of cellular network inference. The fundamental question for DREAM is simple: How can researchers assess how well they are describing the networks of interacting molecules that underlie biological systems? The answer is not so simple. Researchers have used a variety of algorithms to deduce the structure of very different biological and artificial networks, and evaluated their success using various metrics,” write co-organizers Gustavo Stolovitzky of IBM Research, and Andrea Califano from Columbia University.
“What DREAM aims to achieve,” they say, “is a fair comparison of the strengths and weaknesses of the methods and a clear sense of the reliability of the network models they produce.”
Conference sponsors include Columbia University Center for Multiscale Analysis Genomic and Cellular Networks (MAGNet); NIH Roadmap Initiative; IBM Computational Biology Center; The New York Academy of Sciences; and Merck Research Laboratories.
Results must be submitted by October 15. DREAM organizers plan to notify participants of their scores and ranks by
November 15. Rules and contest registration is available at http://wiki.c2b2.columbia.edu/dream/index.php/The_DREAM_Project.
Subscribe to Bio-IT World magazine.