US 12,462,895 B2
T-cell receptor repertoire selection prediction with physical model augmented pseudo-labeling for personalized medicine decision making
Renqiang Min, Princeton, NJ (US); Hans Peter Graf, South Amboy, NJ (US); Erik Kruus, Hillsborough, NJ (US); and Yiren Jian, West Lebanon, NH (US)
Assigned to NEC Corporation, Tokyo (JP)
Filed by NEC Laboratories America, Inc., Princeton, NJ (US)
Filed on Oct. 20, 2022, as Appl. No. 17/969,883.
Claims priority of provisional application 63/307,649, filed on Feb. 8, 2022.
Claims priority of provisional application 63/270,257, filed on Oct. 21, 2021.
Prior Publication US 2023/0129568 A1, Apr. 27, 2023
Int. Cl. G16B 15/00 (2019.01); G06N 3/08 (2023.01); G16B 30/10 (2019.01); G16B 40/20 (2019.01)
CPC G16B 15/00 (2019.02) [G06N 3/08 (2013.01); G16B 30/10 (2019.02); G16B 40/20 (2019.02)] 20 Claims
OG exemplary drawing
 
1. A method for predicting T-Cell receptor (TCR)-peptide interaction, comprising:
training a deep learning model for the predicting TCR-peptide interaction, the training comprising:
determining a multiple sequence alignment (MSA) for a plurality of TCR-peptide pair sequences from a dataset of TCR-peptide pair sequences using a sequence analyzer;
building TCR structures and peptide structures using the MSA and corresponding structures from a Protein Data Bank (PDB) using a MODELLER;
generating an extended TCR-peptide training dataset based on docking energy scores determined by docking peptides to TCRs using physical modeling based on the TCR structures and peptide structures built using the MODELLER; and
classifying and labeling TCR-peptide pairs as positive or negative pairs using pseudo-labels based on the docking energy scores; and
iteratively retraining the deep learning model based on the extended TCR-peptide training dataset and the pseudo-labels until convergence.