Hierarchical multilabel classifier for gene ontology annotation using multihead and multiend deep CNN model

Abstract

Gene ontology annotation is known to be a very complicated multilabel classification task, and the hierarchical multilabel classification (HMC) approaches with local classifiers have been shown to be effective for the task. In a traditional HMC method, a set of hierarchically organized simple local classifiers are usually used, each of which for one hierarchical level separately. In this paper, we propose a novel hierarchical multilabel classifier implementing the whole set of hierarchically organized local classifiers in one deep convolution neural network (CNN) model with multiple heads and multiple ends (MHME). The proposed MHME CNN model consists of three parts: the body part of a deep CNN model shared by different local classifiers for feature extraction and feature mapping; the multiend part of a set of autoencoders performing feature fusion transforming the input vectors of different local classifiers to feature vectors with the same length so as to share the feature mapping part; and the multihead part of a set of linear multilabel classifiers. Furthermore, a sophisticated recursive algorithm is designed to train the MHME CNN model to realize the functions of a set of hierarchically organized local classifiers. In this way, by sharing a deep CNN with multiple local classifiers, we are able to construct more powerful local classifiers for each level with limited training samples, and to achieve better classification performance. Experiment results on various benchmark datasets show that the proposed deep CNN‐based model has better performance than the state‐of‐the‐art traditional models. Moreover, it gives rather good performance even under a transfer learning. © 2020 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

Learning locality-sensitive bucketing functions

“…Our work also extends to various other domains such as predicting homologous protein sequences. Convolutional neural networks have been shown to be effective in protein sequence feature extraction ( Yuan et al 2020 ). Instead of training on edit distance, the model can be easily modified to use the scores of protein sequences as its target for homologous prediction.…”

Deep Protein Subcellular Localization Predictor Enhanced with Transfer Learning of GO Annotation

“…The GO annotation predictor and the subcellular localization predictor share a deep CNN feature extractor, but they work on a pre-training and fine-tuning manner. The GO annotation predictor is a hierarchical multi-label model [22][23][24], consisting of a deep CNN feature extractor and a set of linear multi-label classifiers [25]. It is trained by using a large amount of available experimental GO annotations from various related species.…”

Constructing a PPI Network Based on Deep Transfer Learning for Protein Complex Detection

“…The deep PPI predictor consists of a semi‐supervised SVM classifier and a deep CNN feature extractor. The deep CNN feature extractor is first pretrained in a group of GO annotation and subcellular localization predictors implemented by a multi‐head and multi‐end (MHME) deep CNN model [10–12], using datasets from the type species, then fine‐tuned in a binary PPI detector using experimentally identified PPI samples. In this way, the unknown PPIs are predicted by a deep PPI predictor enhanced with the transfer learning of GO and SL annotations.…”