Machine learning identifies abnormal Ca(2+) transients in human induced pluripotent stem cell-derived cardiomyocytes.
Machine learning identifies abnormal Ca(2+) transients in human induced pluripotent stem cell-derived cardiomyocytes.
复制标题
DOI：
10.1038/s41598-020-73801-x
复制
发表时间：
2020-10-12
影响因子：
4.6
通讯作者：
Xu C
中科院分区：
文献类型：
--
作者：
Hwang H;Liu R;Maxwell JT;Yang J;Xu C
人诱导多能干细胞衍生的心肌细胞（hiPSC-CM）为潜在的临床和研究应用提供了极好的平台。识别异常的Ca 2+瞬变对于评估心肌细胞功能至关重要，这需要劳动密集型的手动工作。因此，我们开发了一个自动评估Ca 2+瞬态异常的分析管道，采用先进的机器学习方法和分析算法。首先，我们采用现有的分析算法来识别Ca 2+瞬态峰，并根据量化的峰特征确定峰异常。其次，我们训练一个峰值级支持向量机（SVM）分类器，使用人类专家评估的峰值异常作为结果和轮廓峰值变量作为预测功能。第三，我们通过使用人类专家对细胞异常的评估作为结果并将量化的细胞水平变量作为预测特征来训练另一个细胞水平SVM分类器。该细胞水平SVM分类器可用于评估额外的Ca 2+瞬态信号。通过将此管道应用于我们的Ca 2+瞬态数据，我们使用200个细胞作为训练数据训练了细胞级SVM分类器，然后在54个细胞的独立数据集中测试了其准确性。结果，我们获得了88%的训练准确率和87%的测试准确率。此外，我们还提供了一个免费的R包来实现我们的高通量CM Ca 2+分析管道。
Human-induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) provide an excellent platform for potential clinical and research applications. Identifying abnormal Ca2+ transients is crucial for evaluating cardiomyocyte function that requires labor-intensive manual effort. Therefore, we develop an analytical pipeline for automatic assessment of Ca2+ transient abnormality, by employing advanced machine learning methods together with an Analytical Algorithm. First, we adapt an existing Analytical Algorithm to identify Ca2+ transient peaks and determine peak abnormality based on quantified peak characteristics. Second, we train a peak-level Support Vector Machine (SVM) classifier by using human-expert assessment of peak abnormality as outcome and profiled peak variables as predictive features. Third, we train another cell-level SVM classifier by using human-expert assessment of cell abnormality as outcome and quantified cell-level variables as predictive features. This cell-level SVM classifier can be used to assess additional Ca2+ transient signals. By applying this pipeline to our Ca2+ transient data, we trained a cell-level SVM classifier using 200 cells as training data, then tested its accuracy in an independent dataset of 54 cells. As a result, we obtained 88% training accuracy and 87% test accuracy. Further, we provide a free R package to implement our pipeline for high-throughput CM Ca2+ analysis.
登录
查看更多内容
影响因子：
23.9
作者：
Burridge, Paul W.;Keller, Gordon;Gold, Joseph D.;Wu, Joseph C.
通讯作者：
Wu, Joseph C.
影响因子：
4.6
作者：
Jha R;Wu Q;Singh M;Preininger MK;Han P;Ding G;Cho HC;Jo H;Maher KO;Wagner MB;Xu C
通讯作者：
Xu C
影响因子：
37.8
作者：
Kitani, Tomoya;Ong, Sang-Ging;Wu, Joseph C.
通讯作者：
Wu, Joseph C.
影响因子：
64.8
作者：
Itzhaki, Ilanit;Maizels, Leonid;Gepstein, Lior
通讯作者：
Gepstein, Lior
影响因子：
20.1
作者：
Sharma A;Marceau C;Hamaguchi R;Burridge PW;Rajarajan K;Churko JM;Wu H;Sallam KI;Matsa E;Sturzu AC;Che Y;Ebert A;Diecke S;Liang P;Red-Horse K;Carette JE;Wu SM;Wu JC
通讯作者：
Wu JC