A number of supervised machine learning models have recently been introduced for the prediction of drug–target interactions based on chemical structure and genomic sequence information. Although these models could offer improved means for many network pharmacology applications, such as repositioning of drugs for new therapeutic uses, the prediction models are often being constructed and evaluated under overly simplified settings that do not reflect the real-life problem in practical applications. Using quantitative drug–target bioactivity assays for kinase inhibitors, as well as a popular benchmarking data set of binary drug–target interactions for enzyme, ion channel, nuclear receptor and G protein-coupled receptor targets, we illustrate here the effects of four factors that may lead to dramatic differences in the prediction results: (i) problem formulation (standard binary classification or more realistic regression formulation), (ii) evaluation data set (drug and target families in the application use case), (iii) evaluation procedure (simple or nested cross-validation) and (iv) experimental setting (whether training and test sets share common drugs and targets, only drugs or targets or neither). Each of these factors should be taken into consideration to avoid reporting overoptimistic drug–target interaction prediction results. We also suggest guidelines on how to make the supervised drug–target interaction prediction studies more realistic in terms of such model formulations and evaluation setups that better address the inherent complexity of the prediction task in the practical applications, as well as novel benchmarking data sets that capture the continuous nature of the drug–target interactions for kinase inhibitors.

Recent studies suggest that exercise alters the gut microbiome. We determined whether six-weeks endurance exercise, without changing diet, affected the gut metagenome and systemic metabolites of overweight women. Previously sedentary overweight women (n = 19) underwent a six-weeks endurance exercise intervention, but two were excluded due to antibiotic therapy. The gut microbiota composition and functions were analyzed by 16S rRNA gene amplicon sequencing and metagenomics. Body composition was analyzed with DXA X-ray densitometer and serum metabolomics with NMR metabolomics. Total energy and energy-yielding nutrient intakes were analyzed from food records using Micro-Nutrica software. Serum clinical variables were determined with KONELAB instrument. Soluble Vascular Adhesion Protein 1 (VAP-1) was measured with ELISA and its' enzymatic activity as produced hydrogen peroxide. The exercise intervention was effective, as maximal power and maximum rate of oxygen consumption increased while android fat mass decreased. No changes in diet were observed. Metagenomic analysis revealed taxonomic shifts including an increase in Akkermansia and a decrease in Proteobacteria. These changes were independent of age, weight, fat % as well as energy and fiber intake. Training slightly increased Jaccard distance of genus level β-diversity. Training did not alter the enriched metagenomic pathways, which, according to Bray Curtis dissimilarity analysis, may have been due to that only half of the subjects' microbiomes responded considerably to exercise. Nevertheless, tranining decreased the abundance of several genes including those related to fructose and amino acid metabolism. These metagenomic changes, however, were not translated into major systemic metabolic changes as only two metabolites, phospholipids and cholesterol in large VLDL particles, decreased after exercise. Training also decreased the amine oxidase activity of pro-inflammatory VAP-1, whereas no changes in CRP were detected. All clinical blood variables were within normal range, yet exercise slightly increased glucose and decreased LDL and HDL. In conclusion, exercise training modified the gut microbiome without greatly affecting systemic metabolites or body composition. Based on our data and existing literature, we propose that especially Akkermansia and Proteobacteria are exercise-responsive taxa. Our results warrant the need for further studies in larger cohorts to determine whether exercise types other than endurance exercise also modify the gut metagenome.

No abstract