Repository Issue Activity (beta)

#689 predict covalent bond using smiles

#566 Boltz-2 docking with metal cofactor

#396 Question about Offline Model Weight Loading and Feature Extraction for Custom Datasets

#575 Large structure prediction failing: Error "WARNING: ran out of memory, skipping batch"

#541 Problems about structure prediction with Templates

#626 remove pinning of dependencies?

#182 Trunk sampling for increased diversity

#653 CPU runs produce distorted structures with excessively long bonds and colliding atoms

#662 CPU inference produces lower quality predictions due to device-specific autocast bug in boltz2.py

#687 training code for boltz-2

#686 Training preprocessing: rcsb.py does not correctly populate cluster_id or msa_id for custom mmCIF datasets

#491 Feature Request: Add option to check current version of installed Boltz

#17 protein length vs. GPU memory

#596 Mismatched Keys in the Validation Step in Boltz1

#572 Details about ATLAS experiments?

#609 Unclear how to specify contact constraints in Boltz2

#619 Generate paired MSAs directly from locally generated unpaired MSAs (taxonomy-based pairing)

#680 [Bug] steering_args['fk_steering'] raises TypeError when steering_args is None (diffusionv2.py:323)

#679 [Bug] pairformer_args.v2 missing from checkpoint hyper_parameters → AttentionPairBias.forward() unexpected kwarg 'k_in'

#616 Questions about contact restraint handling

#617 Inconsistent data in ccd.pkl and ccd.rdb

#639 Inserting modified residues that are not inside CCD chemical database

#613 Missing symmetry writing and ccd.rdb in the training dataset prep

#606 MSA pairing algorithm without using the ColabFold server

#583 Constructing MSA feature array removes non-duplicate paired sequences