School of Computer Science and Technology, Beijing Institute of Technology, Beijing 100081, China.
Bio-Computing Research Center, Harbin Institute of Technology, Shenzhen 518055, China.
Bioinformatics. 2022 May 13;38(10):2712-2718. doi: 10.1093/bioinformatics/btac200.
Therapeutic peptide prediction is important for the discovery of efficient therapeutic peptides and drug development. Researchers have developed several computational methods to identify different therapeutic peptide types. However, these computational methods focus on identifying some specific types of therapeutic peptides, failing to predict the comprehensive types of therapeutic peptides. Moreover, it is still challenging to utilize different properties to predict the therapeutic peptides.
In this study, an adaptive multi-view based on the tensor learning framework TPpred-ATMV is proposed for predicting different types of therapeutic peptides. TPpred-ATMV constructs the class and probability information based on various sequence features. We constructed the latent subspace among the multi-view features and constructed an auto-weighted multi-view tensor learning model to utilize the high correlation based on the multi-view features. Experimental results showed that the TPpred-ATMV is better than or highly comparable with the other state-of-the-art methods for predicting eight types of therapeutic peptides.
The code of TPpred-ATMV is accessed at: https://github.com/cokeyk/TPpred-ATMV.
Supplementary data are available at Bioinformatics online.
治疗性肽预测对于发现高效治疗性肽和药物开发非常重要。研究人员已经开发了几种计算方法来识别不同类型的治疗性肽。然而，这些计算方法侧重于识别某些特定类型的治疗性肽，无法预测治疗性肽的综合类型。此外，利用不同特性来预测治疗性肽仍然具有挑战性。
在这项研究中，我们提出了一种基于张量学习框架的自适应多视图方法 TPpred-ATMV，用于预测不同类型的治疗性肽。TPpred-ATMV 基于各种序列特征构建类别和概率信息。我们构建了多视图特征之间的潜在子空间，并构建了一个自动加权多视图张量学习模型，以利用多视图特征之间的高相关性。实验结果表明，TPpred-ATMV 在预测八种类型的治疗性肽方面优于或与其他最先进的方法高度可比。
TPpred-ATMV 的代码可在 https://github.com/cokeyk/TPpred-ATMV 上访问。
补充数据可在 Bioinformatics 在线获得。
