Multiscale Graph Neural Networks for Protein Residue Contact Map Prediction
Multiscale Graph Neural Networks for Protein Residue Contact Map Prediction
复制标题
DOI：
10.48550/arxiv.2212.02251
复制
发表时间：
2022-12
期刊：
影响因子：
--
通讯作者：
Kuang Liu;R. Kalia;Xinlian Liu;A. Nakano;K. Nomura;P. Vashishta;R. Zamora-Resendiz
中科院分区：
文献类型：
--
作者：
Kuang Liu;R. Kalia;Xinlian Liu;A. Nakano;K. Nomura;P. Vashishta;R. Zamora-Resendiz
机器学习 (ML) 正在彻底改变蛋白质结构分析，其中包括预测蛋白质残基接触图的一个重要子问题，即在给定蛋白质氨基酸序列的情况下，哪些氨基酸残基在空间上非常接近。尽管基于机器学习的蛋白质接触预测最近取得了进展，但预测大范围距离的接触（通常分为短程、中程和长程接触）仍然是一个挑战。在这里，我们从多尺度物理模拟中汲取灵感，提出了一种基于多尺度图神经网络（GNN）的方法，其中涉及循环神经网络（RNN）的标准管道通过三个 GNN 进行了增强，以分别细化短程、中程和长程残留接触的预测能力。 ProteinNet 数据集上的测试结果表明，与传统方法相比，使用所提出的多尺度 RNN+GNN 方法可以提高所有范围接触的准确性，包括最具挑战性的远程接触预测情况。
Machine learning (ML) is revolutionizing protein structural analysis, including an important subproblem of predicting protein residue contact maps, i.e ., which amino-acid residues are in close spatial proximity given the amino-acid sequence of a protein. Despite recent progresses in ML-based protein contact prediction, predicting contacts with a wide range of distances (commonly classified into short-, medium-and long-range contacts) remains a challenge. Here, we propose a multiscale graph neural network (GNN) based approach taking a cue from multiscale physics simulations, in which a standard pipeline involving a recurrent neural network (RNN) is augmented with three GNNs to refine predictive capability for short-, medium- and long-range residue contacts, respectively. Test results on the ProteinNet dataset show improved accuracy for contacts of all ranges using the proposed multiscale RNN+GNN approach over the conventional approach, including the most challenging case of long-range contact prediction.