Shamil Sunyaev's interests run the gamut. One of them is in mutations, which he studies from three different points of view: function, evolution, and medicine — all without picking up a pipette. In his lab at Harvard Medical School, Sunyaev is developing computational and statistical tools to analyze mutations and genetic variations. Also, he is developing bioinformatics tools to deal with the coming sequencing and proteomics data. Sunyaev and his lab are working on methods to predict molecular function from sequence data. For this, Sunyaev takes a comparative genomics approach and collaborates with functional genomics groups. Then, they also take a look at that same data with an eye toward evolutionary genetics and how the mutations they see affect fitness. In particular, Sunyaev is interested in evolutionary patterns and what happens to those variants over time. Finally, he collaborates with medical geneticists to analyze how these genetic variants affect phenotypes. Sunyaev is also preparing for the coming glut of sequence data and the information about rare alleles that will be hidden in these. "Human genome sequencing is becoming so accessible that we will have many, many human genomes," he says. "You have very large amounts of data and sequencing will discover very many variants — and many of them will be rare in the population." In his lab, Sunyaev is working on developing methods to analyze those sequences and plan for upcoming resequencing projects. "Right now, in this regard, we're trying to do population simulations…to look at the available data and try to learn population genetic models and then simulate very large resequencing studies," Sunyaev says. This will help indicate "what strategies are going to be best — what is possible, what is not possible to do, and so forth, to inform future studies and to develop tools to be used for these studies," he adds. Another objective in the Sunyaev lab is developing computational methods for proteomics. One such project has Sunyaev comparing protein interactions and protein complexes across yeast species. Another is part of the SysCode consortium which has the goal of engineering mammalian organs, but in a way that is informed by developmental proteomics. Sunyaev says the challenges facing his lab include incorporating all the different disciplines they cover. "There's so many nice projects, I think we just do too much," he says. Looking ahead The field is moving very quickly, Sunyaev says. "This is what I'm banking on: I think that we will have a lot of genomic data and something will come out of that or not come out [depending] on how it is interpreted," he says. In the 1960s through the 1980s, Sunyaev says, there were a lot of nice analyses developed but there was no data for them. Soon, he says, "the position will be reversed." He expects that there will be an abundance of data, and the bottleneck will be interpreting it all. Publications of note A 2001 paper of Sunyaev's in Human Molecular Genetics estimated that a single human genome contains 1,000 deleterious mutations by examining the effects that amino acid replacements have on protein structure and function. In a more recent paper in the American Journal of Human Genetics, Sunyaev and his colleagues studied the role of low-frequency genetic variants in disease and found that 70 percent of low-frequency missense alleles are mildly deleterious, meaning that they are associated with a loss in fitness. And the Nobel goes to ... Sunyaev would like to win for "a model completely explaining the inheritance of complex traits and the molecular architecture of the traits."
A Better Approach to Interpretation
Premium