October 7, 2008
As the leaders of the Human Genome Project pondered the prospect of full-scale DNA sequencing in the mid 1990s, one of the primary topics of conversation was automation. Could the existing DNA sequencing technology be automated on an industrial scale to churn out vast amounts of data? Now a decade later, that same question has arisen on the next level of the biological realm, proteins. Is it possible to automate on a large scale the analysis of a protein's structure to predict its biological function?
Although the answer remains very much a work in progress, NIDCR-supported scientists and colleagues have developed a progressive - and potentially more efficient - strategy. As published in the September issue of PLoS Computational Biology, the approach pushes aside existing functional definition schemes to take a more quantitative look at the role of existing amino acid residues in protein function. As the authors explain, "To calculate the quantitative values of each residue, we used a combined approach, the metafunctional signature (MFS), which takes into account the individual scores from various function prediction algorithms and generates a composite score for each amino acid residue in a given protein." The MFS is a composite of four biological factors: sequence conservation, evolutionary conservation, structural stability, and amino acid type. The scientists note that in their preliminary work, MFS scores outperform other algorithms that track a single category of information.