.png){kind=icon}
返回
Efficient computation by molecular competition networks
.png){kind=icon}
OA
AI
DOI：10.1103/PhysRevResearch.6.033208.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Most biomolecular systems exhibit computation abilities, which are often achieved through complex networks such as signal transduction networks. Particularly, molecular competition in these networks can introduce crosstalk and serve as a hidden layer for cellular information processing. Despite the increasing evidence of competition contributing to efficient cellular computation, how this occurs and the extent of computational capacity it confers remain elusive. In this study, we introduced a mathematical model for molecular competition networks (MCNs) and employed a machine learning-based optimization method to explore their computational capacity. Our findings revealed that MCNs, when compared to their noncompetitive counterparts, demonstrate superior performance in both discrete decision-making and analog computation tasks. Furthermore, we highlighted the nonnegligible role of weak interactions and limited amounts of resources and examined how biological constraints influence the computational capacity of MCNs. The study suggested the potential of MCNs as efficient computational structures and provided new insights into cellular information processing.
Keyword:
SPECIFICITY
LOGIC
PROMISCUITY
TITRATION
期刊
IF:
4.2
论文数:
7.6K
被引数:
2.7W
机构
引用论文
Finding gene network topologies for given biological function with recurrent neural network.png)
NATURE COMMUNICATIONS
OAAI
NATURE COMMUNICATIONSIF15.7
OAAI分享收藏分享收藏Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations物理信息神经网络: 一种用于解决涉及非线性偏微分方程的正反问题的深度学习框架OAAI
OAAI分享收藏
查看更多内容