Sequence-Based Protein Function Prediction Using Graph Transformers and Deep Representative Learning (1/3)

Project: A - Government Institutionb - National Science and Technology Council

Project Details


This study establishes a novel framework that can improve the interpretation of protein sequences and the prediction of their functions using AlphaFold2-predicted structures, deep representative learning, and graph transformer. The developed framework will be applied to solve different problems of protein function problems (i.e., protein subcellular localization, gene ontology-based functions, or general functions) that have been proven to be the key mechanisms in molecular biology. This is the first study that observes this combination to protein function prediction in particular and bioinformatics sequences in general. For biological insights, we suggest that our method could provide useful information for biologists studying protein sequence patterns or pathogenic mechanisms of mutations, and chemists interested in targeted drug design.
Effective start/end date8/1/227/31/23


  • computational biology
  • natural language processing
  • deep learning
  • graph transformers
  • protein function prediction
  • sequence analysis
  • neural networks
  • AlphaFold2
  • protein structure
  • transport protein
  • Gene Ontology
  • molecular function
  • precision medicine
  • drug design


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.