Eran Segal: Computing expression

Photo: Ahikam Seri

Eran Segal followed a meandering route to the field of computational biology. He began by earning a bachelor's degree in computer science from Tel-Aviv University in 1998, and went on to study in Stanford University's computer science department under Daphne Koller. He also studied genetics at Stanford, where he began to explore how probabilistic models can answer biologic questions.

As a graduate student Segal focused on designing computational models of gene expression. One included a method for identifying groups of coregulated genes and their regulators. ¹ He tested this method on yeast gene data and then applied it to transcription factor Ypl230w, protein kinase Kin82, and the phosphatase Ppt1, to predict regulation functions, targets, and conditions, which he experimentally verified. Koller says that this research joined a relatively new movement to produce testable hypotheses about regulatory relationships that could be worked out in the lab.

Segal also worked on a model that identified 22,163 pairs of genes that are coexpressed in the DNA of humans, flies, worms, and yeast. ² Segal showed that these genes often have a shared function, which would allow researchers to predict the unknown functions of genes based on their coexpressed partners. While comparative genomics has been around for a while, says Koller, "we were the first to do this for gene-expression data across species."

After leaving Stanford in 2004, Segal spent a year as a postdoctoral fellow at Rockefeller University, where he began collaboration with Jon Widom, now at Northwestern University. He returned to Israel in 2005 and joined the department of computer science and applied mathematics at The Weizmann Institute of Science in Rehovot, where he is now a senior scientist.

At Stanford, Segal would create computational models and then look for biologic applications for them. "Now, I find a biological problem and find the algorithm for tackling that problem," he says.

One such problem was the organization of DNA around nucleosomes. In 2006, Segal and Widom hypothesized that the instructions for wrapping DNA around nucleosomes are contained in the DNA itself. Using nucleosome-bound sequences from yeast, they built a model to predict the genome-wide placement of nucleosomes. When applied, they found that this model could explain 50% of in vivo nucleosome positions, ³ a finding that earned Segal the 2007 Overton Prize from the International Society for Computational Biology.

Segal says he's now found the right balance between computer science and biology: In February 2007, he started his own molecular biology lab at Weizmann. "What is special about him," says Widom, "is his incredibly broad foundation of knowledge - from geek biology and geek physical chemistry to very sophisticated computer science."