The five-year project will use genetics to fill in gaps in the story of human history. Scientists at 10 centers around the world are collecting DNA samples from indigenous populations. Scientists from IBM’s Computational Biology Center will search for new patterns and connections within the genetic data. The genetic information will reveal missing details about global human migration and provide new insights into the interconnections of the human species. The resulting public database will house one of the largest collections of human population genetic information ever assembled as a resource for geneticists, historians, and anthropologists. The genetic data, however, are not disease-oriented or clinically relevant. “The genetic markers are confined to the Y chromosome and mitochondrial DNA and give information about ancestry,” says Dr. Royyuru.
In May 2006, IBM joined with WHO, CDC, and other public health institutions worldwide to control the spread of infectious diseases through the Global Pandemic Initiative. IBM will contribute open-source software programs to communities to share information about disease outbreaks and predict how diseases will spread. An epidemiological modeling framework, called Spatio-Temporal Epidemiological Modeller (STEM), will help public health experts and government planners simulate the spread of a disease based on spatial and time factors, such as weather conditions, bird migration routes, and travel patterns.
For instance, if the H5N1 strain of bird flu contaminates commercial poultry flocks as recently occurred in England, STEM could simulate what-if scenarios, such as the impact of placing a community under quarantine or stopping all air flights. “We hope that STEM will become a baseline tool,” says Dr. Royyuru.
A related program, called Project Checkmate, teams IBM researchers with those at The Scripps Research Institute. The Blue Gene supercomputer will model and predict how the influenza virus, including the infamous 1918 strain, mutates over time. “If you know the variations before the virus mutates, you can prepare with vaccines and therapeutics ahead of time,” Dr. Royyuru says. Likely viral variants can be analyzed to determine their potency, whether they escape immune surveillance, and what drugs or vaccines counteract them. “We can do computational modeling of variants on a much larger scale than you can do in a laboratory,” Dr. Royyuru says.