Cardiovascular diseases continue to be the leading cause of death world-26wide. Therefore, understanding the pathogenesis of these diseases, apply-ing it to clinical practice and identifying new therapeutic targets are important for world health. To this end, we are conducting research to elucidate the pre-cise genetic mechanisms underlying these diseases and to promote the clinical applications of genomic information in clinical practice, using cutting-edge technologies such as whole-genome sequencing and machine learning in addi-tion to statistical genetics.Among these cardiovascular diseases, our team mainly targets not only common diseases such as atherosclerotic diseases, arrhythmias and heart fail-ure, but also rare diseases such as Kawasaki disease, chronic thromboembolic pulmonary hypertension and cancer treatment-related cardiac dysfunction. At present, we are: 1) Conducting large-scale studies to understand the genetic underpinnings of ischemic heart disease, the most common atherosclerotic dis-ease, and atrial fibrillation, the most common arrhythmia, as well as identifying the genetic differences between Japanese and Europeans, in collaboration with international consortia. 2) Developing and validating a new genetic analysis method based on a machine learning algorithm that solves the “P greater than N” scenario, where the sample size is small but the number of variants to be analyzed is large. 3) Elucidating of the mechanism of rare cardiovascular dis-eases using human omics data from multi-center patients in Japan. 4) Perform-ing prospective cohort studies to examine the possibility of clinical application of genomic information. 5) Developing a functional analysis system using massively parallel in vitro assays with artificial intelligence. 6) Developing and validating deep learning-based cardiovascular age for clinical applications. We also play an important role in genomic analyses of AMED GRIFIN and AMED intractable disease projects for cardiovascular diseases.Our ultimate goal is to provide better genome-informed diagnostic/manage-ment/treatment approaches to patients suffering from cardiovascular diseases and to medical professionals fighting on the front lines of clinical practice.Figure: Deep learning-based cardiovascular age for clinical applicationsThe chest X-ray (CXR) dataset was randomly divided into train-ing, validation and test datasets. Our deep neural network (DNN) models were trained to estimate age and sex using the training dataset. The weights of the models were initialized with pre-trained weights on ImageNet data and trained using transfer learning and fine-tuning techniques. Various models with different architectures were separately trained. Validation data were only used to tune the hyperparameters and to select the final model. The accuracy of the deep learning model was estimated using a hold-out test dataset. The independent dataset was also used to estimate the performance to verify the generalizability of the trained DNN in an independent popula-tion. The trained DNN was applied to CXRs of heart failure patients to evaluate the association between the estimated age (X-ray age) and various clinical parameters and the clinical outcomes of heart failure.Recent Major PublicationsPatel PN, Ito K, Willcox JAL, Haghighi A, Jang MY, Gorham JM, DePalma SR, Lam L, McDonough B, Johnson R, Lakdawala NK, Roberts A, Barton PJR, Cook SA, Fatkin D, Seidman CE, Seidman JG. Contribution of Noncanonical Splice Variants to TTN Trun-cating Variant Cardiomyopathy. Circ Genom Precis Med 14, e003389 (2021)Hartiala JA, Han Y, Jia Q, Hilser JR, Huang P, Gukasyan J, Schwartzman WS, Cai Z, Biswas S, Trégouët DA, Smith NL; INVENT Consortium; CHARGE Consortium Hemostasis Working Group; GENIUS-CHD Consortium, Seldin M, Pan C, Mehrabian M, Lusis AJ, Bazeley P, Sun YV, Liu C, Quyyumi AA, Scholz M, Thiery J, Delgado GE, Kleber ME, März W, Howe LJ, Asselbergs FW, van Vugt M, Vlachojannis GJ, ... Ito K, Koyama S, Kamatani Y, Komuro I; Biobank Japan, Stolze LK, Romanoski CE, Khan MD, Turner AW, Miller CL, Aherrahrou R, Civelek M, Ma L, Björkegren JLM, Kumar SR, Tang WHW, Hazen SL, Allayee H. Genome-wide analysis identifies novel susceptibility loci for myocardial infarction. Eur Heart J 42, 919 (2021)Miyazawa K, Ito K. The Evolving Story in the Genetic Analysis for Heart Failure. Front Cardiovasc Med 8, 646816 (2021)Invited presentationsIto K. Frontiers of Genomic Research on Lifestyle-related Diseases “Frontiers of Genomic Analysis of Cardiovascular Dis-eases” The 66th Annual Meeting of the Japan Society of Human Genetics and the 28th Annual Meeting of the Japanese Society for Gene Diagnosis and Therapy - Joint Conference 2021 (Yoko-hama/Online) October 2021Ito K. Frontiers of Heart Failure Genomic Medicine “Heart Fail-ure Research as a ‘Common’ Disease from a Genomic Perspec-tive” The 25th Annual Scientific Meeting of the Japanese Heart Failure Society (Online) October 2021Ito K. “How Should We Apply Omics on Cardiovascular Diseases to Clinical Medicine?” CVCT (Cardiovascular Clinical Trialist) - Asia-Pacific Forum (Online) July 2021Ito K. The Promise and Peril of Next-Generation sequencing “Interpreting the Mendelian Genetic Disease Variant of Un-known Significance Using AI” The 67th Annual Meeting of the Japanese Heart Rhythm Society (Online) July 2021Ito K. Genetic Cardiovascular Disease - From Mechanism to Phenotype and GWAS “Complex Genetic Architecture of Coro-nary Artery Disease Revealed by GWAS and Machine Learning” The 85th Annual Scientific Meeting of the Japanese Circulation Society (Yokohama/Online) March 2021Laboratory for Cardiovascular Genomics and InformaticsTeam Leader: Kaoru Ito
元のページ ../index.html#32