jglaser/binding_affinity
A dataset to fine-tune language models on protein-ligand binding affinity prediction.
parse contact matrices from PDBbind and store in chunks
remove hack
fix version number
eliminate index
convert dtype to float32
update notebooks
update datasets
fix typo
add covalent inhibitors split
add dataset with maximum sequence length 512
really remove all KRAS proteins
fix skip download feature
enable offline mode
fix dataset loading script
add to README
update README
update to most recent DBs, add split without KRAS
add IC50 values back
update dataset to only inclulde Ki/Kd (but not IC50/EC50)
make smiles canonical
update deepcoy DUDE predictions
update predictions for D-COID and deepcoy/DEKOIS
update .gitattributes
add deepcoy DEKOIS and DUDE predictions
add RF2 scoring for testing
add D-COID data set
update predictions
add DUDE predictions
split DUD-E into smaller partitions
repartition DUD-E data
add dude processing scripts
add DUDE test data set
add slurm scripts
update notebooks
update loader script
normalize affinities
remove outliers
update requirements.txt
add label
update README
fix typo
fix formatting
amend README.md
fix features
use pyarrow
try fix download
try fix download
tweak script
use all data for training for now
try a different URL
