👤 Hiroshi Mamitsuka

MOTIVATION: With the rapid increase of biomedical articles, large-scale automatic Medical Subject Headings (MeSH) indexing has become increasingly important. FullMeSH, the only method for large-scale MeSH indexing with full text, suffers from three major drawbacks: FullMeSH (i) uses Learning To Rank, which is time-consuming, (ii) can capture some pre-defined sections only in full text and (iii) ignores the whole MEDLINE database. RESULTS: We propose a computationally lighter, full text and deep-learning-based MeSH indexing method, BERTMeSH, which is flexible for section organization in full text. BERTMeSH has two technologies: (i) the state-of-the-art pre-trained deep contextual representation, Bidirectional Encoder Representations from Transformers (BERT), which makes BERTMeSH capture deep semantics of full text. (ii) A transfer learning strategy for using both full text in PubMed Central (PMC) and title and abstract (only and no full text) in MEDLINE, to take advantages of both. In our experiments, BERTMeSH was pre-trained with 3 million MEDLINE citations and trained on ∼1.5 million full texts in PMC. BERTMeSH outperformed various cutting-edge baselines. For example, for 20 K test articles of PMC, BERTMeSH achieved a Micro F-measure of 69.2%, which was 6.3% higher than FullMeSH with the difference being statistically significant. Also prediction of 20 K test articles needed 5 min by BERTMeSH, while it took more than 10 h by FullMeSH, proving the computational efficiency of BERTMeSH. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online. Show less

The task of drug-target interaction prediction holds significant importance in pharmacology and therapeutic drug design. In this paper, we present FRnet-DTI, an auto-encoder based feature manipulation and a convolutional neural network based classifier for drug target interaction prediction. Two convolutional neural networks are proposed: FRnet-Encode and FRnet-Predict. Here, one model is used for feature manipulation and the other one for classification. Using the first method FRnet-Encode, we generate 4096 features for each of the instances in each of the datasets and use the second method, FRnet-Predict, to identify interaction probability employing those features. We have tested our method on four gold standard datasets extensively used by other researchers. Experimental results shows that our method significantly improves over the state-of-the-art method on three out of four drug-target interaction gold standard datasets on both area under curve for Receiver Operating Characteristic (auROC) and area under Precision Recall curve (auPR) metric. We also introduce twenty new potential drug-target pairs for interaction based on high prediction scores. The source codes and implementation details of our methods are available from https://github.com/farshidrayhanuiu/FRnet-DTI/ and also readily available to use as an web application from http://farshidrayhan.pythonanywhere.com/FRnet-DTI/ . Show less

The US National Library of Medicine (NLM) uses the Medical Subject Headings (MeSH) (see Note 1 ) to index almost all 24 million citations in MEDLINE, which greatly facilitates the application of biomedical information retrieval and text mining. Large-scale automatic MeSH indexing has two challenging aspects: the MeSH side and citation side. For the MeSH side, each citation is annotated by only 12 (on average) out of all 28,000 MeSH terms. For the citation side, all existing methods, including Medical Text Indexer (MTI) by NLM, deal with text by bag-of-words, which cannot capture semantic and context-dependent information well. To solve these two challenges, we developed the MeSHLabeler and DeepMeSH. By utilizing "learning to rank" (LTR) framework, MeSHLabeler integrates multiple types of information to solve the challenge in the MeSH side, while DeepMeSH integrates deep semantic representation to solve the challenge in the citation side. MeSHLabeler achieved the first place in both BioASQ2 and BioASQ3, and DeepMeSH achieved the first place in both BioASQ4 and BioASQ5 challenges. DeepMeSH is available at http://datamining-iip.fudan.edu.cn/deepmesh . Show less

Identifying drug-target interactions is crucial for the success of drug discovery. Approaches based on machine learning for this problem can be divided into two types: feature-based and similarity-based methods. By utilizing the "Learning to rank" framework, we propose a new method, DrugE-Rank, to combine these two different types of methods for improving the prediction performance of new candidate drugs and targets. DrugE-Rank is available at http://datamining-iip.fudan.edu.cn/service/DrugE-Rank/ . Show less

Computationally predicting drug-target interactions is useful to select possible drug (or target) candidates for further biochemical verification. We focus on machine learning-based approaches, particularly similarity-based methods that use drug and target similarities, which show relationships among drugs and those among targets, respectively. These two similarities represent two emerging concepts, the chemical space and the genomic space. Typically, the methods combine these two types of similarities to generate models for predicting new drug-target interactions. This process is also closely related to a lot of work in pharmacogenomics or chemical biology that attempt to understand the relationships between the chemical and genomic spaces. This background makes the similarity-based approaches attractive and promising. This article reviews the similarity-based machine learning methods for predicting drug-target interactions, which are state-of-the-art and have aroused great interest in bioinformatics. We describe each of these methods briefly, and empirically compare these methods under a uniform experimental setting to explore their advantages and limitations. Show less