A machine learning approach to predict cellular mechanical stresses in response to chemical perturbation
A machine learning approach to predict cellular mechanical stresses in response to chemical perturbation
复制标题
一种机器学习方法来预测响应化学扰动的细胞机械应力
DOI：
10.1016/j.bpj.2023.07.016
复制
发表时间：
2023
影响因子：
3.4
通讯作者：
Steward, R.L.
中科院分区：
文献类型：
--
作者：
SubramanianBalachandar, VigneshAravind;Islam, Md. Mydul;Steward, R.L.
在细胞-细胞水平和细胞-基质水平上产生的机械应力已经被认为在许多生理和病理过程中是重要的。然而，各种化合物对上述机械应力的影响知之甚少，阻碍了新疗法的发现，并成为该领域的障碍。为了克服这一障碍，我们实施了两种方法：1）单层边界预测器和2）离散窗口预测器，其利用逐步线性回归或二次支持向量机机器学习模型来预测牵引和细胞间应力对化学扰动的剂量依赖性响应。我们使用从受0.2或2μg/mL药物浓度沿着的样品收集的实验牵引和细胞间应力数据以及从亮场图像提取的细胞形态学特性作为预测因子来训练我们的模型。为了证明我们的机器学习模型的预测能力，我们预测了响应于0和1μg/mL药物浓度的牵引力和细胞间应力，这些药物浓度在训练集中没有使用。结果显示，离散化窗口预测器仅用四个样本（292张图像）训练，分别使用二次支持向量机和逐步线性回归模型对看不见的样本图像进行最佳预测。
Mechanical stresses generated at the cell-cell level and cell-substrate level have been suggested to be important in a host of physiological and pathological processes. However, the influence various chemical compounds have on the mechanical stresses mentioned above is poorly understood, hindering the discovery of novel therapeutics, and representing a barrier in the field. To overcome this barrier, we implemented two approaches: 1) monolayer boundary predictor and 2) discretized window predictor utilizing either stepwise linear regression or quadratic support vector machine machine learning model to predict the dose-dependent response of tractions and intercellular stresses to chemical perturbation. We used experimental traction and intercellular stress data gathered from samples subject to 0.2 or 2μg/mL drug concentrations along with cell morphological properties extracted from the bright-field images as predictors to train our model. To demonstrate the predictive capability of our machine learning models, we predicted tractions and intercellular stresses in response to 0 and 1μg/mL drug concentrations which were not utilized in the training sets. Results revealed the discretized window predictor trained just with four samples (292 images) to best predict both intercellular stresses and tractions using the quadratic support vector machine and stepwise linear regression models, respectively, for the unseen sample images.