DeepGANnel: Synthesis of fully annotated single molecule patch-clamp data using generative adversarial networks.
DeepGANnel: Synthesis of fully annotated single molecule patch-clamp data using generative adversarial networks.
复制标题
DOI：
10.1371/journal.pone.0267452
复制
发表时间：
2022
期刊：
影响因子：
3.7
通讯作者：
Barrett-Jolley R
中科院分区：
文献类型：
--
作者：
Ball STM;Celik N;Sayari E;Abdul Kadir L;O'Brien F;Barrett-Jolley R
由于缺乏可用的训练数据，用于“单离子通道”记录的自动分析工具的开发受到阻碍。对于基于机器学习的工具，需要非常大的训练集以及逐个样本点标记的数据（例如，每 100 微秒 1 个样本点）。在实验环境中，此类数据被标记为人工监督，虽然这对于简单的实验分析是可行的，但使用手工制作的大数据方法生成所需的大量数据集是不可行的。在这项工作中，我们的目标是开发生成模拟离子通道数据的方法，该数据不受噪声和潜在隐马尔可夫模型的假设和先验知识的影响。我们成功地利用生成对抗网络（GAN）构建了一个端到端管道，用于从小型带注释的离子通道“种子”记录中生成无限量的标记训练数据，并且这不需要理论动态离子通道特性的先验知识。我们的方法利用 2D CNN 来维持原始记录和理想记录之间的同步时间关系。我们通过 5 个不同的数据源证明了该方法的适用性，并通过真实数据和合成数据之间的 t-SNE 和 UMAP 投影比较显示了真实性。该模型可以轻松扩展到需要并行标记的其他时间序列数据，例如标记的心电图信号或原始纳米孔测序数据。
Development of automated analysis tools for “single ion channel” recording is hampered by the lack of available training data. For machine learning based tools, very large training sets are necessary with sample-by-sample point labelled data (e.g., 1 sample point every 100microsecond). In an experimental context, such data are labelled with human supervision, and whilst this is feasible for simple experimental analysis, it is infeasible to generate the enormous datasets that would be necessary for a big data approach using hand crafting. In this work we aimed to develop methods to generate simulated ion channel data that is free from assumptions and prior knowledge of noise and underlying hidden Markov models. We successfully leverage generative adversarial networks (GANs) to build an end-to-end pipeline for generating an unlimited amount of labelled training data from a small, annotated ion channel “seed” record, and this needs no prior knowledge of theoretical dynamical ion channel properties. Our method utilises 2D CNNs to maintain the synchronised temporal relationship between the raw and idealised record. We demonstrate the applicability of the method with 5 different data sources and show authenticity with t-SNE and UMAP projection comparisons between real and synthetic data. The model would be easily extendable to other time series data requiring parallel labelling, such as labelled ECG signals or raw nanopore sequencing data.
登录
查看更多内容
影响因子：
3.7
作者：
Imbrici P;Camerino DC;Tricarico D
通讯作者：
Tricarico D
DOI：
10.1142/s1793048013300053
发表时间：
2013-12-01
期刊：
Biophysical Reviews and Letters
影响因子：
--
作者：
Nicolai, Christopher;Sachs, Frederick
通讯作者：
Sachs, Frederick
影响因子：
5.6
作者：
Mobasheri, Ali;Lewis, Rebecca;Maxwell, Judith E. J.;Hill, Claire;Womack, Matthew;Barrett-Jolley, Richard
通讯作者：
Barrett-Jolley, Richard
影响因子：
64.8
作者：
NEHER, E;SAKMANN, B
通讯作者：
SAKMANN, B
影响因子：
1.2
作者：
Anderson, David F.;Ermentrout, Bard;Thomas, Peter J.
通讯作者：
Thomas, Peter J.