Scholarlink

We study graph neural network learning for transcriptomics with limited amount of labeled data. Our study reveals that simple GNN architectures perform well and do not suffer from over-fitting as the more sophisticated ones. Our study shows that although contrastive learning as a pretraining strategy has been successful in predicting properties such as formation and binding energy, it is not effective for transcriptomics. We propose attention pooling as an effective measure for improving the prediction performance. It beats all other models in our experiments. Track:CSCI-RTAI

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Study on Contrastive Graph Neural Network Pretraining for Predicting Transcriptome Profiles

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jiaji Ma,
  • Scott Auerbach,
  • Guojing Cong

摘要

We study graph neural network learning for transcriptomics with limited amount of labeled data. Our study reveals that simple GNN architectures perform well and do not suffer from over-fitting as the more sophisticated ones. Our study shows that although contrastive learning as a pretraining strategy has been successful in predicting properties such as formation and binding energy, it is not effective for transcriptomics. We propose attention pooling as an effective measure for improving the prediction performance. It beats all other models in our experiments. Track:CSCI-RTAI