A Novel Computational Method for Deriving Protein Secondary Structure Topologies Using Cryo-EM Density Maps and Multiple Secondary Structure Predictions

A key idea in de novo secondary structure topology determination methods is to calculate an optimal mapping between the observed secondary structure traces in a Cryo-EM density image and the predicted secondary structures on the protein sequence. The problem becomes much more complicated in presence of multiple secondary structure predictions for the protein sequence (for example those predicted by different prediction methods). We present a novel computational method that elegantly and efficiently solves the problem of dealing with multiple secondary structure predictions and calculating the optimal mapping. The proposed method uses a two-step approach - it first uses the consensus positions of the secondary structures to produce top K topologies, and then it uses a dynamic programming method to find the optimal placement for the secondary structure traces of the density image. The method was tested using twelve proteins of three types. We observed that the rank of the true topologies is consistently improved with the use of multiple secondary structure predictions over single prediction. The results show that the algorithm is robust and works well even in presence of errors/misses in predicted secondary structures from the image or the sequence. The results also show that the algorithm is efficient and is able to handle proteins with as many as thirty-three helices.