Toward an accurate prediction of inter-residue distances in proteins using 2D recursive neural networks

By Predrag Kukic, Claudio Mirabello, Giuseppe Tradigo, Ian Walsh, Pierangelo Veltri, Gianluca Pollastri. 2014

48 citations recorded by the source corpus.

Protein inter-residue contact maps provide a translation and rotation invariant topological representation of a protein. They can be used as an intermediary step in protein structure predictions. However, the prediction of contact maps represents an unbalanced problem as far fewer examples of contacts than non-contacts exist in a protein structure. In this study we explore the possibility of completely eliminating the unbalanced nature of the contact map prediction problem by predicting real-value distances between residues. Predicting full inter-residue distance maps and applying them in protein structure predictions has been relatively unexplored in the past. We initially demonstrate that the use of native-like distance maps is able to reproduce 3D structures almost identical to the targets, giving an average RMSD of 0.5Å. In addition, the corrupted physical maps with an introduced random error of ±6Å are able to reconstruct the targets within an average RMSD of 2Å. After demonstrating the reconstruction potential of distance maps, we develop two classes of predictors using two-dimensional recursive neural networks: an ab initio predictor that relies only on the protein sequence and evolutionary information, and a template-based predictor in which additional structural homology information is provided. We find that the ab initio predictor is able to reproduce distances with an RMSD of 6Å, regardless of the evolutionary content provided. Furthermore, we show that the template-based predictor exploits both sequence and structure information even in cases of dubious homology and outperforms the best template hit with a clear margin of up to 3.7Å. Lastly, we demonstrate the ability of the two predictors to reconstruct the CASP9 targets shorter than 200 residues producing the results similar to the state of the machine learning art approach implemented in the Distill server. The methodology presented here, if complemented by more complex reconstruction protocols, can represent a possible path to improve machine learning algorithms for 3D protein structure prediction. Moreover, it can be used as an intermediary step in protein structure predictions either on its own or complemented by NMR restraints.

Subfields: Computational Theory and Mathematics, Materials Chemistry, Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Materials Chemistry

In Molecular Biology

Citations in the corpus

Distance-based protein folding powered by deep learning (2018, methodology): “Kukic et al. 26 proposed a recursive neural network method for inter-residue distance prediction and studied distance-based protein folding, but the predicted distance has a large error and the resultant 3D models have poor quality.”
De novo Protein Structure Prediction by Coupling Contact With Distance Profile (2020, background): “Since Miyazawa and Jernigan proposed the concept of residue-residue potentials [30], many researchers begin to predict the distance between residues [31], [32], [33].”
Deep learning methods in protein structure prediction (2020, methodology): “Both ab initio and template-based predictors have been developed to predict maps (as well as 1D PSA) [29,96].”
Ultrafast end-to-end protein structure prediction enables high-throughput exploration of uncharacterized proteins (2022, unknown): “Our approach constructs a directly learned representation of the sequences in an MSA, starting from a one-hot encoding of the sequences.”