Uncertainty quantified discovery of chemical reaction systems via Bayesian scientific machine learning.

IF 2.3
Frontiers in systems biology Pub Date : 2024-03-08 eCollection Date: 2024-01-01 DOI:10.3389/fsysb.2024.1338518
Emily Nieves, Raj Dandekar, Chris Rackauckas
{"title":"Uncertainty quantified discovery of chemical reaction systems via Bayesian scientific machine learning.","authors":"Emily Nieves, Raj Dandekar, Chris Rackauckas","doi":"10.3389/fsysb.2024.1338518","DOIUrl":null,"url":null,"abstract":"<p><p>The recently proposed Chemical Reaction Neural Network (CRNN) discovers chemical reaction pathways from time resolved species concentration data in a deterministic manner. Since the weights and biases of a CRNN are physically interpretable, the CRNN acts as a digital twin of a classical chemical reaction network. In this study, we employ a Bayesian inference analysis coupled with neural ordinary differential equations (ODEs) on this digital twin to discover chemical reaction pathways in a probabilistic manner. This allows for estimation of the uncertainty surrounding the learned reaction network. To achieve this, we propose an algorithm which combines neural ODEs with a preconditioned stochastic gradient langevin descent (pSGLD) Bayesian framework, and ultimately performs posterior sampling on the neural network weights. We demonstrate the successful implementation of this algorithm on several reaction systems by not only recovering the chemical reaction pathways but also estimating the uncertainty in our predictions. We compare the results of the pSGLD with that of the standard SGLD and show that this optimizer more efficiently and accurately estimates the posterior of the reaction network parameters. Additionally, we demonstrate how the embedding of scientific knowledge improves extrapolation accuracy by comparing results to purely data-driven machine learning methods. Together, this provides a new framework for robust, autonomous Bayesian inference on unknown or complex chemical and biological reaction systems.</p>","PeriodicalId":73109,"journal":{"name":"Frontiers in systems biology","volume":"4 ","pages":"1338518"},"PeriodicalIF":2.3000,"publicationDate":"2024-03-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12341973/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Frontiers in systems biology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.3389/fsysb.2024.1338518","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/1/1 0:00:00","PubModel":"eCollection","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

The recently proposed Chemical Reaction Neural Network (CRNN) discovers chemical reaction pathways from time resolved species concentration data in a deterministic manner. Since the weights and biases of a CRNN are physically interpretable, the CRNN acts as a digital twin of a classical chemical reaction network. In this study, we employ a Bayesian inference analysis coupled with neural ordinary differential equations (ODEs) on this digital twin to discover chemical reaction pathways in a probabilistic manner. This allows for estimation of the uncertainty surrounding the learned reaction network. To achieve this, we propose an algorithm which combines neural ODEs with a preconditioned stochastic gradient langevin descent (pSGLD) Bayesian framework, and ultimately performs posterior sampling on the neural network weights. We demonstrate the successful implementation of this algorithm on several reaction systems by not only recovering the chemical reaction pathways but also estimating the uncertainty in our predictions. We compare the results of the pSGLD with that of the standard SGLD and show that this optimizer more efficiently and accurately estimates the posterior of the reaction network parameters. Additionally, we demonstrate how the embedding of scientific knowledge improves extrapolation accuracy by comparing results to purely data-driven machine learning methods. Together, this provides a new framework for robust, autonomous Bayesian inference on unknown or complex chemical and biological reaction systems.

不确定性量化发现化学反应系统通过贝叶斯科学机器学习。
最近提出的化学反应神经网络(CRNN)以确定性的方式从时间分辨的物种浓度数据中发现化学反应路径。由于CRNN的权重和偏差在物理上是可解释的,因此CRNN充当经典化学反应网络的数字孪生。在这项研究中,我们采用贝叶斯推理分析结合神经常微分方程(ode)对这个数字双胞胎以概率方式发现化学反应途径。这允许对学习反应网络周围的不确定性进行估计。为此,我们提出了一种将神经ode与预条件随机梯度朗格万下降(pSGLD)贝叶斯框架相结合的算法,并最终对神经网络权值进行后验抽样。我们证明了该算法在几个反应系统上的成功实现,不仅恢复了化学反应途径,而且估计了我们预测中的不确定性。我们将pSGLD的结果与标准SGLD的结果进行了比较,结果表明该优化器更有效、更准确地估计了反应网络参数的后验。此外,我们通过将结果与纯数据驱动的机器学习方法进行比较,展示了科学知识的嵌入如何提高外推的准确性。总之,这为未知或复杂的化学和生物反应系统的鲁棒,自主贝叶斯推理提供了一个新的框架。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信