Supervised Machine Learning for Classification of the Electrophysiological Effects of Chronotropic Drugs on Human Induced Pluripotent Stem Cell-Derived Cardiomyocytes.
Supervised Machine Learning for Classification of the Electrophysiological Effects of Chronotropic Drugs on Human Induced Pluripotent Stem Cell-Derived Cardiomyocytes.
复制标题
DOI：
10.1371/journal.pone.0144572
复制
发表时间：
2015
期刊：
影响因子：
3.7
通讯作者：
Gratton E
中科院分区：
文献类型：
--
作者：
Heylman C;Datta R;Sobrino A;George S;Gratton E
有监督的机器学习可用于预测人类心肌细胞暴露于哪些药物。使用从已知暴露于不同药物的人类心肌细胞收集的电生理数据，可以训练监督机器学习算法来识别和分类已经暴露于未知药物的细胞。此外，学习算法提供关于每个数据参数对总体分类的相对贡献的信息。每个分类的准确性的概率和置信度也可以由算法确定。在这项研究中，β-肾上腺素能药物，普萘洛尔和异丙肾上腺素，对来自人诱导多能干细胞（hiPS-CM）的心肌细胞的电生理作用进行了评估。电生理数据收集使用高时间分辨率的双光子显微镜的电压敏感染料作为报告的膜电压。结果表明，我们的算法能够准确地评估，分类和预测hiPS-CM膜去极化暴露于变时性药物。
Supervised machine learning can be used to predict which drugs human cardiomyocytes have been exposed to. Using electrophysiological data collected from human cardiomyocytes with known exposure to different drugs, a supervised machine learning algorithm can be trained to recognize and classify cells that have been exposed to an unknown drug. Furthermore, the learning algorithm provides information on the relative contribution of each data parameter to the overall classification. Probabilities and confidence in the accuracy of each classification may also be determined by the algorithm. In this study, the electrophysiological effects of β–adrenergic drugs, propranolol and isoproterenol, on cardiomyocytes derived from human induced pluripotent stem cells (hiPS-CM) were assessed. The electrophysiological data were collected using high temporal resolution 2-photon microscopy of voltage sensitive dyes as a reporter of membrane voltage. The results demonstrate the ability of our algorithm to accurately assess, classify, and predict hiPS-CM membrane depolarization following exposure to chronotropic drugs.