Evotuning protocols for Transformer-based variant effect prediction on multi-domain proteins
Evotuning protocols for Transformer-based variant effect prediction on multi-domain proteins
复制标题
基于 Transformer 的多域蛋白变异效应预测的 Evotuning 协议
DOI：
10.1101/2021.03.05.434175
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Saito Yutaka
中科院分区：
文献类型：
--
作者：
Yamaguchi Hideki;Saito Yutaka
准确的变异效应预测对蛋白质工程具有广泛的影响。为此目的，最近的机器学习方法是基于表示学习，通过表示学习，从未标记的序列中学习和生成特征向量。然而，目前还不清楚如何有效地从同源序列中学习工程靶蛋白的进化特性，考虑到蛋白质的序列水平结构，称为结构域架构（DA）。此外，没有最佳的协议建立将这些属性到Transformer，众所周知的神经网络在自然语言处理研究中表现最好。本文提出了DA意识的进化微调，或'evotuning'，协议的transformer为基础的变体效应预测，考虑到同源性搜索，微调和序列矢量化策略的各种组合。我们对具有不同功能和DA的各种蛋白质进行了详尽的评估。结果表明，我们的协议取得了显着更好的性能比以前的DA不知道的。注意力地图的可视化表明，结构信息是通过没有直接监督的唤起来整合的，可能会导致更好的预测精度。
Accurate variant effect prediction has broad impacts on protein engineering. Recent machine learning approaches toward this end are based on representation learning, by which feature vectors are learned and generated from unlabeled sequences. However, it is unclear how to effectively learn evolutionary properties of an engineering target protein from homologous sequences, taking into account the protein’s sequence-level structure called domain architecture (DA). Additionally, no optimal protocols are established for incorporating such properties into Transformer, the neural network well-known to perform the best in natural language processing research. This article proposes DA-aware evolutionary fine-tuning, or ‘evotuning’, protocols for Transformer-based variant effect prediction, considering various combinations of homology search, fine-tuning and sequence vectorization strategies. We exhaustively evaluated our protocols on diverse proteins with different functions and DAs. The results indicated that our protocols achieved significantly better performances than previous DA-unaware ones. The visualizations of attention maps suggested that the structural information was incorporated by evotuning without direct supervision, possibly leading to better prediction accuracy.