用于离子通道Markov模型参数优化的进化策略方法

Evolutionary strategy for parameter optimization in Markov model for ion channel

目的针对离子通道Markov建模过程中参数的数量较多、通常很难确定等问题,提出一个以细胞电生理知识为基础用于离子通道Markov建模的进化算法。方法首先,从大量的电压-电流的数据选取若干具有代表性的曲线组合用于建模。然后,从选定的曲线选出若干能够反映曲线变化的特征点用以构建进化策略方法的适应度函数。最后,设计具体的遗传算子并进行模型参数优化。结果对采集自实际细胞的离子通道数据进行了建模及优化,得到的模型对实际数据拟合误差很小。结论通过分析离子通道Markov模型及数据,设计了一套以进化算法为基础的建模优化方法,建立的模型能很好地说明通道的电生理特性。

Objective There are too many parameters in Markov modeling for ion channel. This paper proposes an evolutionary algorithm （EA） method based on cell electrophysiological knowledge to solve the problem. Methods Firstly, certain typical voltage-current curves were selected t-ore a large number of voltage- current data as the training curves of the evolutionary strategy （ES）. Then, feature infm^nation was extracted from the selected curves. Finally, the genetic operators of the ES were designed to optimize the model parameters. Results The data collected from the real experiments were modeled and optimized, and the model could fit the real data very well. Conclusions A modeling and optimization method is designed based on evolutionary algorithm with the analysis on Markov model and data of ion channel. The model can reflect ion channel＇s electrophysiologieal characteristics exactly.