Accelerating Antibody Informatics with Machine Learning

Time: 5:00 pm
day: Day One


  • CDR-H3 is a fingerprint region of each antibody, and its conformation is key for antigen recognition
  • Examining various kinds of machine learning methods (Random Forest, Decision Tree, Light GBM, SVM, MLP) for the prediction of CDR-H3’s structure type
  • Recurrent neural network and convolutional neural network in combination with reinforcement learning for the de novo design of developable antibody sequences