The active gene annotation corpus (AGAC) was developed to support knowledge discovery for drug repurposing. The AGAC track of the BioNLP Open Shared Tasks 2019 was organized, to facilitate cross-disciplinary collaboration across BioNLP and Pharmacoinformatics communities, for drug repurposing. The AGAC track consists of three subtasks: 1) named entity recognition, 2) thematic relation extraction, and 3) loss of function (LOF) / gain of function (GOF) topic classification. The AGAC track was participated by five teams, of which the performance is compared and analyzed. The results revealed a substantial room for improvement in the design of the task, which we analyzed in terms of "imbalanced data", "selective annotation" and "latent topic annotation".
Off-target effects played a vital role in the pharmacological understanding of drug efficacy and this research aimed to use text mining strategy to curate molecular level information and unveil the mechanism of off-target effect caused by the usage of anti-multiple myeloma (MM) drugs. After training a hybrid CNN-CRF-LSTM neural network upon the training data from TAC 2017 benchmark database, we extracted all of the side effects of 16 anti-MM drugs from drug labels, and combined the results with existed database. Afterwards, gene targets of anti-MM drugs were obtained by using structure similarity, and their related phenotypes were retrieved from Human Phenotype Ontology. Furthermore, linked phenotypes to candidate genes and adverse reaction of known drugs formed a knowledge graph. Through regulation analysis upon intersected phenotypes of drugs and target genes, an off-target effect caused by SLC7A7 was found, which with high possibility unveiled the pharmacological mechanism of side effect after using combination of anti-MM drugs.
The table-based fact verification task has recently gained widespread attention and yet remains to be a very challenging problem. It inherently requires informative reasoning over natural language together with different numerical and logical reasoning on tables (e.g., count, superlative, comparative). Considering that, we exploit mixture-of-experts and present in this paper a new method: Self-adaptive Mixtureof-Experts Network (SaMoE). Specifically, we have developed a mixture-of-experts neural network to recognize and execute different types of reasoning-the network is composed of multiple experts, each handling a specific part of the semantics for reasoning, whereas a management module is applied to decide the contribution of each expert network to the verification result. A self-adaptive method is developed to teach the management module combining results of different experts more efficiently without external knowledge. The experimental results illustrate that our framework achieves 85.1% accuracy on the benchmark dataset TAB-FACT, comparable with the previous state-ofthe-art models. We hope our framework can serve as a new baseline for table-based verification. Our code is available at https: //github.com/THUMLP/SaMoE.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.