Detail

Prediction of RNA Secondary Structure

During a long time , RNA was believed to be a passive intermediate in genetic expression. Since a few decades however, a series of RNA families are being discovered which play a role similar to proteins, thus calling for structural analysis. I will focus on these RNAs and target my presentation on the difficulties that may appear when trying to predict their (secondary) structure.

When only a few RNA homologues are available, one of the most famous approach is a dynamic programming scheme based on a set of recursions proposed by Sankoff in 1985, which may be seen as a combination of the classical recursions for sequence alignment and for single sequence secondary structure prediction. Its main drawback lies in its time and space complexities, that although polynomial are prohibitive in practice. Hence a number of heuristics were developed by some authors to face that difficulty and turn the recursions more or less usable with natural sequences. In the talk we present a new strategy for the design of heuristics that drastically reduce time and space requirement together with initial experimental results.