据介绍，药物安全倡议已认可人iPSC来源心肌细胞（hiPSC-CM）作为预测药物诱导心律失常的体外模型。然而，人类定义的体外心律失常特征在多大程度上预测了实际的临床风险，目前一直备受争议。
研究人员训练了一个卷积神经网络分类器（CNN），来学习与致命性Torsade de Pointes心律失常相关的hiPSC-CM体外动作电位记录特征。CNN分类器准确预测了药物引起的人心律失常风险。来自不同健康供体的hiPSC-CM试验药物的风险状况相似。相比之下，导致患者心律失常性心肌病的致病突变显著增加了hiPSC-CM对某些中高风险药物的致心律失常倾向。
因此，深度学习可以识别与临床心律失常相关的体外心律失常特征，并识别患者遗传学对药物性心律失常风险的影响。
附：英文原文
Title: A deep learning platform to assess drug proarrhythmia risk
Author: Ricardo Serrano, Dries A.M. Feyen, Arne A.N. Bruyneel, Anna P. Hnatiuk, Michelle M. Vu, Prashila L. Amatya, Isaac Perea-Gil, Maricela Prado, Timon Seeger, Joseph C. Wu, Ioannis Karakikes, Mark Mercola
Issue&Volume: 2022-12-22
Abstract: Drug safety initiatives have endorsed human iPSC-derived cardiomyocytes (hiPSC-CMs)as an in vitro model for predicting drug-induced cardiac arrhythmia. However, the extent to whichhuman-defined features of in vitro arrhythmia predict actual clinical risk has been much debated. Here, we trained aconvolutional neural network classifier (CNN) to learn features of in vitro action potential recordings of hiPSC-CMs that are associated with lethal Torsadede Pointes arrhythmia. The CNN classifier accurately predicted the risk of drug-inducedarrhythmia in people. The risk profile of the test drugs was similar across hiPSC-CMsderived from different healthy donors. In contrast, pathogenic mutations that causearrhythmogenic cardiomyopathies in patients significantly increased the proarrhythmicpropensity to certain intermediate and high-risk drugs in the hiPSC-CMs. Thus, deeplearning can identify in vitro arrhythmic features that correlate with clinical arrhythmia and discern the influenceof patient genetics on the risk of drug-induced arrhythmia.
DOI: 10.1016/j.stem.2022.12.002
Source: https://www.cell.com/cell-stem-cell/fulltext/S1934-5909(22)00486-6
Cell Stem Cell：《细胞—干细胞》，创刊于2007年。隶属于细胞出版社，最新IF：25.269
官方网址：https://www.cell.com/cell-stem-cell/home
投稿链接：https://www.editorialmanager.com/cell-stem-cell/default.aspx
本期文章：《细胞—干细胞》：Online/在线发表
