A Robust Drug-Target Interaction Prediction Framework with Capsule Network and Transfer Learning.

Drug-target interactions (DTIs) are considered a crucial component of drug design and drug discovery. To date, many computational methods were developed for drug-target interactions, but they are insufficiently informative for accurately predicting DTIs due to the lack of experimentally verified negative datasets, inaccurate molecular feature representation, and ineffective DTI classifiers. Therefore, we address the limitations of randomly selecting negative DTI data from unknown drug-target pairs by establishing two experimentally validated datasets and propose a capsule network-based framework called CapBM-DTI to capture hierarchical relationships of drugs and targets, which adopts pre-trained bidirectional encoder representations from transformers (BERT) for contextual sequence feature extraction from target proteins through transfer learning and the message-passing neural network (MPNN) for the 2-D graph feature extraction of compounds to accurately and robustly identify drug-target interactions. We compared the performance of CapBM-DTI with state-of-the-art methods using four experimentally validated DTI datasets of different sizes, including human (Homo sapiens) and worm (Caenorhabditis elegans) species datasets, as well as three subsets (new compounds, new proteins, and new pairs). Our results demonstrate that the proposed model achieved robust performance and powerful generalization ability in all experiments. The case study on treating COVID-19 demonstrates the applicability of the model in virtual screening.

Compared the Microbiota Profiles between Samples from Bronchoalveolar Lavage and Endotracheal Aspirates in Severe Pneumonia: A Real-World Experience.

Lower respiratory tract sampling from endotracheal aspirate (EA) and bronchoalveolar lavage (BAL) are both common methods to identify pathogens in severe pneumonia. However, the difference between these two methods in microbiota profiles remains unclear. We compared the microbiota profiles of pairwise EA and BAL samples in ICU patients with respiratory failure due to severe pneumonia. We prospectively enrolled 50 ICU patients with new onset of pneumonia requiring mechanical ventilation. EA and BAL were performed on the first ICU day, and samples were analyzed for microbial community composition via 16S rRNA metagenomic sequencing. Pathogens were identified in culture medium from BAL samples in 21 (42%) out of 50 patients. No difference was observed in the antibiotic prescription pattern, ICU mortality, or hospital mortality between BAL-positive and BAL-negative patients. The microbiota profiles in the EA and BAL samples are similar with respect to diversity, microbial composition, and microbial community correlations. The antibiotic treatment regimen was rarely changed based on the BAL findings. The samples from BAL did not provide more information than EA in the microbiota profiles. We suggest that EA is more useful than BAL for microbiome identification in mechanically ventilated patients.

Vaginal microbiome variances in sample groups categorized by clinical criteria of bacterial vaginosis.

BACKGROUND: One of the most common and recurrent vaginal infections is bacterial vaginosis (BV). The diagnosis is based on changes to the "normal" vaginal microbiome; however, the normal microbiome appears to differ according to reproductive status and ethnicity, and even among individuals within these groups. The Amsel criteria and Nugent score test are widely used for diagnosing BV; however, these tests are based on different criteria, and so may indicate distinct changes in the vaginal microbial community. Nevertheless, few studies have compared the results of these test against metagenomics analysis. METHODS: Vaginal flora samples from 77 participants were classified according to the Amsel criteria and Nugent score test. The microbiota composition was analyzed using 16S ribosome RNA gene amplicon sequencing. Bioinformatics analysis and multivariate statistical analysis were used to evaluate the microbial diversity and function. RESULTS: Only 3 % of the participants diagnosed BV negative using the Amsel criteria (A-) were BV-positive according to the Nugent score test (N+), while over half of the BV-positive patients using the Amsel criteria (A+) were BV-negative according to the Nugent score test (N-). Thirteen genera showed significant differences in distribution among BV status defined by BV tests (e.g., A - N-, A + N- and A + N+). Variations in the four most abundant taxa, Lactobacillus, Gardnerella, Prevotella, and Escherichia, were responsible for most of this dissimilarity. Furthermore, vaginal microbial diversity differed significantly among the three groups classified by the Nugent score test (N-, N+, and intermediate flora), but not between the Amsel criteria groups. Numerous predictive microbial functions, such as bacterial chemotaxis and bacterial invasion of epithelial cells, differed significantly among multiple BV test, but not between the A- and A+ groups. CONCLUSIONS: Metagenomics analysis can greatly expand our current understanding of vaginal microbial diversity in health and disease. Metagenomics profiling may also provide more reliable diagnostic criteria for BV testing.

miRTarBase update 2018: a resource for experimentally validated microRNA-target interactions.

MicroRNAs (miRNAs) are small non-coding RNAs of ∼ 22 nucleotides that are involved in negative regulation of mRNA at the post-transcriptional level. Previously, we developed miRTarBase which provides information about experimentally validated miRNA-target interactions (MTIs). Here, we describe an updated database containing 422 517 curated MTIs from 4076 miRNAs and 23 054 target genes collected from over 8500 articles. The number of MTIs curated by strong evidence has increased ∼1.4-fold since the last update in 2016. In this updated version, target sites validated by reporter assay that are available in the literature can be downloaded. The target site sequence can extract new features for analysis via a machine learning approach which can help to evaluate the performance of miRNA-target prediction tools. Furthermore, different ways of browsing enhance user browsing specific MTIs. With these improvements, miRTarBase serves as more comprehensively annotated, experimentally validated miRNA-target interactions databases in the field of miRNA related research. miRTarBase is available at http://miRTarBase.mbc.nctu.edu.tw/.