DISTEMA: distance map-based estimation of single protein model accuracy with attentive 2D convolutional neural network

Abstract

Estimation of the accuracy (quality) of protein structural models is important for both prediction and use of protein structural models. Deep learning methods have been used to integrate protein structure features to predict the quality of protein models. Inter-residue distances are key information for predicting protein's tertiary structures and therefore have good potentials to predict the quality of protein structural models. However, few methods have been developed to fully take advantage of predicted inter-residue distance maps to estimate the accuracy of a single protein structural model. We developed an attentive 2D convolutional neural network (CNN) with channel-wise attention to take only a raw difference map between the inter-residue distance map calculated from a single protein model and the distance map predicted from the protein sequence as input to predict the quality of the model. The network comprises multiple convolutional layers, batch normalization layers, dense layers, and Squeeze-and-Excitation blocks with attention to automatically extract features relevant to protein model quality from the raw input without using any expert-curated features. We evaluated DISTEMA's capability of selecting the best models for CASP13 targets in terms of ranking loss of GDT-TS score. The ranking loss of DISTEMA is 0.079, lower than several state-of-the-art single-model quality assessment methods. The work demonstrates that using raw inter-residue distance information alone with deep learning can predict the quality of protein structural models reasonably well.

A Gated Graph Transformer for Protein Complex Structure Quality Assessment and its Performance in CASP15

“…Previously, proteins have been represented as tableau data in the form of hand-crafted features [20]. Along this line, many works [21, 22] have represented proteins using pairwise information embeddings such as residue-residue distance maps and contact maps. Recently, describing proteins as graphs has become a popular means of representing proteins, as such representations can learn and leverage proteins’ geometric information more naturally.…”

EGR: Equivariant Graph Refinement and Assessment of 3D Protein Complex Structures

“…As DL-driven structure prediction methods have matured, new methods for quality assessment (QA) of protein structures have also been developed to facilitate automatic ranking of protein structures. In particular, 2D convolutional neural networks (CNNs) [42], 3D CNNs [43], and GNNs [44,45,9] have recently been adopted for structure ranking.…”

High-Performance Deep Learning Toolbox for Genome-Scale Prediction of Protein Structure and Function

“…We applied the attention mechanism to predict protein residue-residue contact [6]. In a recent study, the DISTEMA method [8] used only the single raw distance map as the input for predicting a decoy structure's quality score. Building on our this work, we inserted self-attention modules into the CNN to catch the global attention information by using the depth-wise separable convolution layer to generate the query key value matrices for self-attention.…”

“…However, training a vision-transformer model demands significant computational power. To train the model, we used CASP8-13 as our training dataset, and we generated a difference map [8] for each decoy as the input feature.…”