Two categories of protein-ligand interaction data, binding affinity and contact prediction, are provided for language model fine-tuning. These labels enable the development of predictive models for biochemical interactions and structural contacts between proteins and ligands.
Use Cases
- Fine-tune language models to predict binding affinity from protein and ligand data.
- Train models to perform contact prediction using the provided labels.
- Evaluate the accuracy of predictive models on protein-ligand binding affinity tasks.
- Develop scoring functions for molecular docking by leveraging the binding affinity and contact prediction features.
Strengths
- Includes binding affinity labels for protein-ligand interaction modeling.
- Features contact prediction data for identifying contacts between proteins and ligands.
- Formatted specifically for language model fine-tuning on structural biology tasks.