Search

The Animal Metabolite Database (AMDB, https://amdb.online) is a freely accessible database with built-in statistical analysis tools, allowing one to browse and compare quantitative metabolomics data and raw NMR and MS data, as well as sample metadata, with a focus on the metabolite concentrations rather than on the raw data itself. AMDB also functions as a platform for the metabolomics community, providing convenient deposition and exchange of quantitative metabolomic data. To date, the majority of the data in AMDB relate to the metabolite content of the eye lens and blood of vertebrates, primarily wild species from Siberia, Russia and laboratory rodents. However, data on other tissues (muscle, heart, liver, brain, and more) are also present, and the list of species and tissues is constantly growing. Typically, every sample in AMDB contains concentrations of 60–90 of the most abundant metabolites, provided in nanomoles per gram of wet tissue weight (nmol/g). We believe that AMDB will become a widely used tool in the community, as typical metabolite baseline concentrations in tissues of animal models will aid in a wide variety of fundamental and applied scientific fields, including, but not limited to, animal modeling of human diseases, assessment of medical formulations, and evolutionary and environmental studies. [ABSTRACT FROM AUTHOR]

Chemical named entity recognition (NER) is an active field of research in biomedical natural language processing. To facilitate the development of new and superior chemical NER systems, BioCreative released the CHEMDNER corpus, an extensive dataset of diverse manually annotated chemical entities. Most of the systems trained on the corpus rely on complicated hand-crafted rules or curated databases for data preprocessing, feature extraction and output post-processing, though modern machine learning algorithms, such as deep neural networks, can automatically design the rules with little to none human intervention. Here we explored this approach by experimenting with various deep learning architectures for targeted tokenisation and named entity recognition. Our final model, based on a combination of convolutional and stateful recurrent neural networks with attention-like loops and hybrid word- and character-level embeddings, reaches near human-level performance on the testing dataset with no manually asserted rules. To make our model easily accessible for standalone use and integration in third-party software, we’ve developed a Python package with a minimalistic user interface. [ABSTRACT FROM AUTHOR]