Prediction of Protein Topologies Using GIOHMMs and GRNNs

We develop and test new machine learning methods for the prediction of topological representations of protein structures in the form of coarse-or ne-grained contact or distance maps that are translation and rotation invariant. The methods are based on generalized input-output hidden Markov models (GIOHMMs) and generalized recursive neural networks (GRNNs). The methods are used to predict topology directly in the ne-grained case and, in the coarse-grained case, indirectly by rst learning how to score candidate graphs and then using the scoring function to search the space of possible conngurations. Computer simulations show that the pre-dictors achieve state-of-the-art performance.