boltz
boltz copied to clipboard
jwohlwend
Official repository for the Boltz biomolecular interaction models
By using the bond type information, the bond order of the ligand can be retained for the PDB writer by changing the CONECT writer section: CONECT 1 2 becomes CONECT...
Getting out of memory errors with Boltz-1 version 0.3.2 when predicting protein with ~2300 residues. My GPU has 48GB memory; my CPU RAM is 256GB. I tried downloading a fresh...
Dear team, thanks for releasing the code of this powerful tool. I tested it on a few complex targets of my interest; the results are very promising. However, it failed...
Fixes GH-181 Check chain name when input data file is parsed and throw a meaningful error message if it's too long.
Hi all, just ran into this error on a hemichannel (6 connexin) system (same as before). I can run this prediction with ColabFold, but not with Boltz-1. YAML Input: ```...
Please add a --version argument. Pip installation does not bring along the pyproject.toml file. On a multi-user system with multiple versions of boltz this would be very helpful.
Hi - is it possible to model covalently bound ligands that are not present in the CCD? Thanks
Hi, I got a segmentation fault with my boltz run for a rather long sequence (>4000AA). My input xml file was this: ``` version: 1 # Optional, defaults to 1...
Hello, I'm currently re-training a Boltz model using the provided datasets. When I attempt to leverage DDP (Distributed Data Parallel) for multi-GPU training, I encounter an error related to the...
When I have a chain ID longer than 5 characters I get: ``` File "/usr/local/lib/python3.11/dist-packages/boltz/main.py", line 613, in predict process_inputs( File "/usr/local/lib/python3.11/dist-packages/lightning_utilities/core/rank_zero.py", line 42, in wrapped_fn return fn(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^...
