Generalizable Machine Learning in Neuroscience Using Graph Neural Networks

被引:11
|
作者
Wang, Paul Y. [1 ,2 ]
Sapra, Sandalika [1 ,3 ]
George, Vivek Kurien [1 ,4 ]
Silva, Gabriel A. [1 ,4 ,5 ]
机构
[1] Univ Calif San Diego, Ctr Engn Nat Intelligence, La Jolla, CA 92093 USA
[2] Univ Calif San Diego, Dept Phys, La Jolla, CA 92093 USA
[3] Univ Calif San Diego, Dept Elect & Comp Engn, La Jolla, CA 92093 USA
[4] Univ Calif San Diego, Dept Bioengn, La Jolla, CA 92093 USA
[5] Univ Calif San Diego, Dept Neurosci, La Jolla, CA 92093 USA
来源
关键词
calcium imaging; graph neural network; deep learning; C elegans; motor action classification;
D O I
10.3389/frai.2021.618372
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Although a number of studies have explored deep learning in neuroscience, the application of these algorithms to neural systems on a microscopic scale, i.e. parameters relevant to lower scales of organization, remains relatively novel. Motivated by advances in whole-brain imaging, we examined the performance of deep learning models on microscopic neural dynamics and resulting emergent behaviors using calcium imaging data from the nematode C. elegans. As one of the only species for which neuron-level dynamics can be recorded, C. elegans serves as the ideal organism for designing and testing models bridging recent advances in deep learning and established concepts in neuroscience. We show that neural networks perform remarkably well on both neuron-level dynamics prediction and behavioral state classification. In addition, we compared the performance of structure agnostic neural networks and graph neural networks to investigate if graph structure can be exploited as a favourable inductive bias. To perform this experiment, we designed a graph neural network which explicitly infers relations between neurons from neural activity and leverages the inferred graph structure during computations. In our experiments, we found that graph neural networks generally outperformed structure agnostic models and excel in generalization on unseen organisms, implying a potential path to generalizable machine learning in neuroscience.
引用
收藏
页数:11
相关论文
共 50 条
  • [41] Canopy classification using LiDAR: a generalizable machine learning approach
    R. Sky Jones
    Racha Elkadiri
    Henrique Momm
    [J]. Modeling Earth Systems and Environment, 2023, 9 : 2371 - 2384
  • [42] Supervised learning of process discovery techniques using graph neural networks
    Sommers, Dominique
    Menkovski, Vlado
    Fahland, Dirk
    [J]. INFORMATION SYSTEMS, 2023, 115
  • [43] A Transfer Approach Using Graph Neural Networks in Deep Reinforcement Learning
    Yang, Tianpei
    You, Heng
    Hao, Jianye
    Zheng, Yan
    Taylor, Matthew E.
    [J]. THIRTY-EIGHTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, VOL 38 NO 15, 2024, : 16352 - 16360
  • [44] Extreme Learning Machine to Graph Convolutional Networks
    Goncalves, Thales
    Nonato, Luis Gustavo
    [J]. INTELLIGENT SYSTEMS, PT II, 2022, 13654 : 601 - 615
  • [45] HiC-GNN: A generalizable model for 3D chromosome reconstruction using graph convolutional neural networks
    Hovenga, Van
    Kalita, Jugal
    Oluwadare, Oluwatosin
    [J]. COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL, 2023, 21 : 812 - 836
  • [46] Graph Transformer: Learning Better Representations for Graph Neural Networks
    Wang, Boyuan
    Cui, Lixin
    Bai, Lu
    Hancock, Edwin R.
    [J]. STRUCTURAL, SYNTACTIC, AND STATISTICAL PATTERN RECOGNITION, S+SSPR 2020, 2021, 12644 : 139 - 149
  • [47] Comprehensive Mapping of Continuous/Switching Circuits in CCM and DCM to Machine Learning Domain Using Homogeneous Graph Neural Networks
    Khamis, Ahmed K.
    Agamy, Mohammed
    [J]. IEEE OPEN JOURNAL OF CIRCUITS AND SYSTEMS, 2023, 4 : 50 - 69
  • [48] Controlling Graph Dynamics with Reinforcement Learning and Graph Neural Networks
    Meirom, Eli A.
    Maron, Haggai
    Mannor, Shie
    Chechik, Gal
    [J]. INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 139, 2021, 139
  • [49] Learning Adaptive Neighborhoods for Graph Neural Networks
    Saha, Avishkar
    Mendez, Oscar
    Russell, Chris
    Bowden, Richard
    [J]. 2023 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2023), 2023, : 22484 - 22493
  • [50] Learning to Count Isomorphisms with Graph Neural Networks
    Yu, Xingtong
    Liu, Zemin
    Fang, Yuan
    Zhang, Xinming
    [J]. THIRTY-SEVENTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, VOL 37 NO 4, 2023, : 4845 - 4853