A, E.; Virko, E.; Kudlak, B.; Fredriksson, R.; Spjuth, O.; Schi h, H.B. Integrating Statistical

A, E.; Virko, E.; Kudlak, B.; Fredriksson, R.; Spjuth, O.; Schi h, H.B. Integrating Statistical and Machine-Learning Strategy for Meta-Analysis of Bisphenol A-Exposure Datasets Reveals Effects on Mouse Gene Expression inside Pathways of Apoptosis and Cell Survival. Int. J. Mol. Sci. 2021, 22, 10785. ten.3390/ ijms221910785 Academic Editors: Ashis Basu and Anthony LemariReceived: 1 September 2021 Accepted: 27 September 2021 Published: 5 October8 7Machine Finding out Applications and Deep Finding out Group, JetBrains Investigation, Kantemirovskaya Str., 2, 197342 St. Petersburg, Russia; elena.kartysheva@jetbrains (E.K.); virkoliza@gmail (E.V.) Department of Neuroscience, Functional Pharmacology, University of Uppsala, BMC, Husargatan three, Box 593, 751 24 Uppsala, Sweden; [email protected] (M.J.W.); [email protected] (H.B.S.) Details Technologies and Programming Faculty, ITMO University, Kronverksky Pr. 49, bldg. A, 197101 St. Petersburg, Russia St. Petersburg College of Physics, Mathematics, and Laptop Science, HSE University, 16 Soyuza Pechatnikov Street, 190121 St. Petersburg, Russia Department of Analytical Chemistry, Faculty of Chemistry, Gdansk University of Technologies, 11/12 Narutowicza Str., 80-233 Gdansk, Poland; [email protected] Division of Pharmaceutical Biosciences, Molecular Neuropharmacology, Uppsala Biomedical Centre, University of Uppsala, Husargatan 3, Box 591, 751 24 Uppsala, Sweden; [email protected] Division of Pharmaceutical Biosciences, Pharmaceutical Bioinformatics, Uppsala Biomedical Centre, University of Uppsala, Husargatan three, Box 591, 751 24 Uppsala, Sweden; [email protected] Institute of Translational Medicine and Biotechnology, I. M. Sechenov Initial Moscow State Healthcare University, Trubetskay Str. eight, bldg 2, 119991 Moscow, Russia Correspondence: nina.lukashina@jetbrainsPublisher’s Note: MDPI stays neutral with regard to jurisdictional claims in Ucf-101 site published maps and institutional affiliations.Abstract: Bisphenols are vital environmental pollutants which can be extensively studied due to diverse detrimental effects, even though the molecular mechanisms behind these effects are less effectively understood. Like other environmental pollutants, bisphenols are being tested in several experimental models, developing big expression datasets located in open access storage. The meta-analysis of such datasets is, however, pretty complicated for numerous causes. Here, we created an integrating statistical and machine-learning model method for the meta-analysis of bisphenol A (BPA) exposure datasets from various mouse tissues. We constructed 3 joint datasets following 3 different techniques for dataset integration: in specific, employing all popular genes in the datasets, uncorrelated, and not co-expressed genes, respectively. By applying machine understanding strategies to these datasets, we identified genes whose expression was significantly affected in all the BPA microanalysis information tested; these involved in the regulation of cell survival consist of: Tnfr2, Hgf-Met, Agtr1a, Bdkrb2; signaling via Mapk8 (Jnk1)); DNA repair (Hgf-Met, Mgmt); apoptosis (Tmbim6, Bcl2, Apaf1); and cellular junctions (F11r, Cldnd1, Ctnd1 and Yes1). Our outcomes highlight the benefit of combining current datasets for the integrated evaluation of a specific subject when person datasets are restricted in size. Key phrases: BPA; BPA-exposure datasets; DNA repair; cellular junctionCopyright: 2021 by the authors. Licensee M.