Biology Department Seminar

In 1990s, it was widely believed that protein folding would soon be solved by the molecular dynamics technique using supercomputers. In 2000s, along with the rapid progress witnessed by Critical Assessments of Techniques for Protein Structure Prediction (CASPs), it was conceivable that protein structure could be predicted reliably in a similar way to a Blast search. However, after decades these seemingly reachable goals are still lingering beyond our ability. So, is computational biology just icing on the cake, as some biologists thought, or a field that is capable of solving difficult problems in biology? In this presentation, two examples will be shown where structural bioinformatics and molecular simulation techniques were combined to study biological complexes. In one case, an atomic model of the entire virus was constructed based on a approx. 5 Angstrom cryoEM map. In another case, a structural model of a channel-receptor domain complex was predicted along with the detailed interface interactions, which were confirmed experimentally at both domain level and full-length protein level.