BindCraft is an open-source and automated pipeline for de novo protein binder design pipeline using AlphaFold2 backpropagation, MPNN, and PyRosetta.
Allocate an interactive session and run the program. Sample session:
[user@biowulf]$ interactive --gres=gpu:v100x:1,lscratch:10 --mem=24g -c16
[user@cn0816 ~]$ module load bindcraft
[+] Loading cuDNN/8.1.0.77/CUDA-11.2.2 libraries...
[+] Loading CUDA Toolkit 11.2.2 ...
[+] Loading PyRosetta 387.py3.10 on cn0816
[+] Loading bindcraft 1.5.0 ...
[user@cn0816 ~]$ cp -r $BC_SRC .
[user@cn0816 ~]$ cd BindCraft-1.5.0
[user@cn0816 ~]$ bindcraft.py -h
vailable GPUs:
Tesla V100-SXM2-32GB1: gpu
usage: bindcraft.py [-h] --settings SETTINGS [--filters FILTERS] [--advanced ADVANCED]
Script to run BindCraft binder design.
options:
-h, --help show this help message and exit
--settings SETTINGS, -s SETTINGS
Path to the basic settings.json file. Required.
--filters FILTERS, -f FILTERS
Path to the filters.json file used to filter design. If not provided, default will be used.
--advanced ADVANCED, -a ADVANCED
Path to the advanced.json file with additional design settings. If not provided, default will
be used.
[user@cn0816 ~]$ ./bindcraft.py \
--settings './settings_target/PDL1.json' \
--filters './settings_filters/default_filters.json' \
--advanced './settings_advanced/default_4stage_multimer.json'
Available GPUs:
Tesla V100-SXM2-32GB1: gpu
┌──────────────────────────────────────────────────────────────────────────────┐
│ PyRosetta-4 │
│ Created in JHU by Sergey Lyskov and PyRosetta Team │
│ (C) Copyright Rosetta Commons Member Institutions │
│ │
│ NOTE: USE OF PyRosetta FOR COMMERCIAL PURPOSES REQUIRE PURCHASE OF A LICENSE │
│ See LICENSE.PyRosetta.md or email license@uw.edu for details │
└──────────────────────────────────────────────────────────────────────────────┘
PyRosetta-4 2024 [Rosetta PyRosetta4.Release.python310.linux 2024.39+release.59628fbc5bc09f1221e1642f1f8d157ce49b1410 2024-09-23T07:49:48] retrieved from: http://www.pyrosetta.org
Running binder design for target PDL1
Design settings used: default_4stage_multimer
Filtering designs based on default_filters
Starting trajectory: PDL1_l86_s981562
Stage 1: Test Logits
1 models [0] recycles 1 hard 0 soft 0.02 temp 1 loss 11.80 helix 1.90 pae 0.81 i_pae 0.80 con 4.63 i_con 4.00 plddt 0.30 ptm 0.55 i_ptm 0.11 rg 10.75
2 models [1] recycles 1 hard 0 soft 0.04 temp 1 loss 8.78 helix 1.08 pae 0.74 i_pae 0.71 con 4.38 i_con 3.74 plddt 0.40 ptm 0.56 i_ptm 0.13 rg 1.70
3 models [3] recycles 1 hard 0 soft 0.05 temp 1 loss 7.98 helix 0.77 pae 0.59 i_pae 0.59 con 3.79 i_con 3.74 plddt 0.47 ptm 0.59 i_ptm 0.21 rg 0.97
4 models [0] recycles 1 hard 0 soft 0.07 temp 1 loss 7.93 helix 0.71 pae 0.56 i_pae 0.61 con 3.47 i_con 3.91 plddt 0.52 ptm 0.57 i_ptm 0.17 rg 1.30
5 models [3] recycles 1 hard 0 soft 0.09 temp 1 loss 6.29 helix 0.71 pae 0.42 i_pae 0.48 con 2.40 i_con 3.33 plddt 0.72 ptm 0.59 i_ptm 0.23 rg 1.62
6 models [3] recycles 1 hard 0 soft 0.11 temp 1 loss 6.23 helix 0.72 pae 0.42 i_pae 0.49 con 2.20 i_con 3.45 plddt 0.75
...
Starting trajectory took: 0 hours, 6 minutes, 3 seconds
Starting trajectory: PDL1_l84_s11317
Stage 1: Test Logits
1 models [0] recycles 1 hard 0 soft 0.02 temp 1 loss 12.96 helix 2.12 pae 0.84 i_pae 0.85 con 4.85 i_con 4.22 plddt 0.29 ptm 0.54 i_ptm 0.08 rg 13.32
2 models [4] recycles 1 hard 0 soft 0.04 temp 1 loss 11.23 helix 0.97 pae 0.75 i_pae 0.73 con 4.72 i_con 3.85 plddt 0.41 ptm 0.56 i_ptm 0.13 rg 8.25
...
Initial trajectory pLDDT good, continuing: 0.88
Stage 1: Additional Logits Optimisation
51 models [1] recycles 1 hard 0 soft 0.04 temp 1 loss 9.82 helix 2.20 pae 0.89 i_pae 0.94 con 4.38 i_con 4.97 plddt 0.21 ptm 0.11 i_ptm 0.05 rg 1.84
52 models [1] recycles 1 hard 0 soft 0.08 temp 1 loss 6.55 helix 1.36 pae 0.45 i_pae 0.45 con 2.83 i_con 3.48 plddt 0.59 ptm 0.61 i_ptm 0.26 rg 1.18
...
Optimised logit trajectory pLDDT: 0.91
Stage 2: Softmax Optimisation
76 models [1] recycles 1 hard 0 soft 1 temp 0.96 loss 6.71 helix 1.00 pae 0.46 i_pae 0.40 con 3.62 i_con 2.83 plddt 0.57 ptm 0.69 i_ptm 0.53 rg 0.92
...
Softmax trajectory pLDDT good, continuing: 0.86
Stage 3: One-hot Optimisation
...
125 models [3] recycles 1 hard 1 soft 1 temp 0.01 loss 3.87 helix 1.51 pae 0.23 i_pae 0.17 con 2.21 i_con 1.66 plddt 0.77 ptm 0.84 i_ptm 0.79 rg 1.04
One-hot trajectory pLDDT good, continuing: 0.78
Stage 4: PSSM Semigreedy Optimisation
Running semigreedy optimization...
140 models [0] recycles 1 hard 1 soft 0 temp 1 loss 3.28 helix 1.42 pae 0.17 i_pae 0.13 con 1.89 i_con 1.44 plddt 0.82 ptm 0.87 i_ptm 0.84 rg 0.88
Trajectory successful, final pLDDT: 0.82
Starting trajectory took: 0 hours, 4 minutes, 27 seconds
Fixing interface residues: B13,B14,B18,B19,B22,B23,B25,B26,B29,B30,B33,B34,B58,B62,B65,B66,B68,B69,B72,B73,B75,B76,B79,B80,B83
Base AF2 filters not passed for PDL1_l84_s11317_mpnn1, skipping interface scoring
Unmet filter conditions for PDL1_l84_s11317_mpnn2
Base AF2 filters not passed for PDL1_l84_s11317_mpnn3, skipping interface scoring
Unmet filter conditions for PDL1_l84_s11317_mpnn4
Base AF2 filters not passed for PDL1_l84_s11317_mpnn5, skipping interface scoring
...
Base AF2 filters not passed for PDL1_l84_s11317_mpnn17, skipping interface scoring
Unmet filter conditions for PDL1_l84_s11317_mpnn18
Base AF2 filters not passed for PDL1_l84_s11317_mpnn19, skipping interface scoring
Base AF2 filters not passed for PDL1_l84_s11317_mpnn20, skipping interface scoring
No accepted MPNN designs found for this trajectory.
Design and validation of trajectory PDL1_l84_s11317 took: 0 hours, 7 minutes, 40 seconds
Starting trajectory: PDL1_l128_s352235
Stage 1: Test Logits
...
jectory successful, final pLDDT: 0.93
Starting trajectory took: 0 hours, 5 minutes, 54 seconds
Fixing interface residues: B21,B22,B24,B25,B28,B29,B31,B32,B33,B35,B36,B37,B39,B48,B115,B118,B119,B122,B123
PDL1_l128_s352235_mpnn1 passed all filters
Base AF2 filters not passed for PDL1_l128_s352235_mpnn2, skipping interface scoring
Base AF2 filters not passed for PDL1_l128_s352235_mpnn3, skipping interface scoring
...
End the interactive session:
[user@cn0816 ~]$ exit salloc.exe: Relinquishing job allocation 46116226 [user@biowulf ~]$