Protein-Peptide Interaction Benchmark Group (+TCR-Epitope)

TDC-2 introduces the Protein-Peptide Binding Affinity prediction task. The predictive, non-generative task is to learn a model estimating a function of a protein, peptide, antigen processing pathway, biological context, and interaction features. It outputs a binding affinity value or binary label indicating strong or weak binding. The binary label can also include additional biomarkers, such as allowing for a positive label if and only if the binding interaction is specific. To account for additional biomarkers beyond binding affinity value, our task is specified with a binary label.
TDC-2 provides datasets and benchmarks for a generalized protein-peptide binding interaction prediction task and a TCR-Epitope binding interaction prediction task.

To access a generalized protein-peptide benchmark in the group, use the following code:

from tdc.benchmark_group.protein_peptide_group import ProteinPeptideGroup
group = ProteinPeptideGroup()
train, val = group.get_train_valid_split() # val dataset will be empty. use the train dataset if fine-tuning desired.
test = group.get_test()

## --- (optional) train your model --- ##

predictions = model.predict(test)  # modify as per your model code and test output
out = group.evaluate(predictions)

To access a TCR-Epitope benchmark in the group, use the following code:

from tdc.benchmark_group.protein_peptide_group import ProteinPeptideGroup
group = TCREpitopeGroup()
train, val = group.get_train_valid_split() 
test = group.get_test()

## --- train your model --- ##

predictions = model.predict(test)  # modify as per your model code and test output
res = group.evaluate(predictions)

Follow the instructions on how to use the BenchmarkGroup class and obtain training, validation, and test sets, and how to submit your model to the leaderboard.

The evaluation metric is AUC. For TCR-Epitope, we provide other metrics as well. See corresponding leaderboards.