ZhiyuanChen commited on
Commit
196edb1
·
verified ·
1 Parent(s): fe22f33

Refine OFoldX model card provenance

Browse files
Files changed (1) hide show
  1. README.md +4 -2
README.md CHANGED
@@ -49,6 +49,7 @@ Converted Boltz-2 structure-prediction checkpoint for biomolecular interaction m
49
 
50
  - **Upstream Project**: Boltz-2
51
  - **Source Checkpoint**: `boltz2_conf_final.ckpt`
 
52
  - **Primary Paper**: [Boltz-2: Towards Accurate and Efficient Binding Affinity Prediction](https://doi.org/10.1101/2025.06.14.659707)
53
  - **Upstream License**: MIT for upstream Boltz code and weights
54
 
@@ -73,6 +74,7 @@ Converted Boltz-2 structure-prediction checkpoint for biomolecular interaction m
73
  - **Hub repository**: [oteam/boltz2](https://huggingface.co/oteam/boltz2)
74
  - **Upstream paper**: [Boltz-2: Towards Accurate and Efficient Binding Affinity Prediction](https://doi.org/10.1101/2025.06.14.659707)
75
  - **Upstream repository**: [Boltz-2](https://github.com/jwohlwend/boltz)
 
76
  - **Code**: [`ofoldx/models/boltz2/model.py`](https://github.com/OTeam-AI4S/OFoldX/tree/main/ofoldx/models/boltz2/model.py)
77
  - **Project repository**: [https://github.com/OTeam-AI4S/OFoldX](https://github.com/OTeam-AI4S/OFoldX)
78
  - **Issues**: [https://github.com/OTeam-AI4S/OFoldX/issues](https://github.com/OTeam-AI4S/OFoldX/issues)
@@ -112,11 +114,11 @@ metadata for loading it.
112
 
113
  ### Training Data
114
 
115
- The upstream Boltz-2 report describes a biomolecular interaction training set with expanded structure data, molecular-dynamics ensembles, self-distilled predictions, and affinity supervision. OFoldX does not redistribute the training set.
116
 
117
  ### Training Procedure
118
 
119
- OFoldX converts the upstream Boltz-2 checkpoint into `model.safetensors` plus an OFoldX manifest; it does not run Boltz-2 training.
120
 
121
  ## Evaluation
122
 
 
49
 
50
  - **Upstream Project**: Boltz-2
51
  - **Source Checkpoint**: `boltz2_conf_final.ckpt`
52
+ - **Source Release**: [https://huggingface.co/boltzgen/boltzgen-1](https://huggingface.co/boltzgen/boltzgen-1)
53
  - **Primary Paper**: [Boltz-2: Towards Accurate and Efficient Binding Affinity Prediction](https://doi.org/10.1101/2025.06.14.659707)
54
  - **Upstream License**: MIT for upstream Boltz code and weights
55
 
 
74
  - **Hub repository**: [oteam/boltz2](https://huggingface.co/oteam/boltz2)
75
  - **Upstream paper**: [Boltz-2: Towards Accurate and Efficient Binding Affinity Prediction](https://doi.org/10.1101/2025.06.14.659707)
76
  - **Upstream repository**: [Boltz-2](https://github.com/jwohlwend/boltz)
77
+ - **Source checkpoint release**: [https://huggingface.co/boltzgen/boltzgen-1](https://huggingface.co/boltzgen/boltzgen-1)
78
  - **Code**: [`ofoldx/models/boltz2/model.py`](https://github.com/OTeam-AI4S/OFoldX/tree/main/ofoldx/models/boltz2/model.py)
79
  - **Project repository**: [https://github.com/OTeam-AI4S/OFoldX](https://github.com/OTeam-AI4S/OFoldX)
80
  - **Issues**: [https://github.com/OTeam-AI4S/OFoldX/issues](https://github.com/OTeam-AI4S/OFoldX/issues)
 
114
 
115
  ### Training Data
116
 
117
+ The Boltz-2 report describes structure data from PDB structures before 2023-06-01, experimental NMR collections, MD trajectory datasets, AF2 monomer distillation, Boltz-1 complex distillation, and affinity data from PubChem, ChEMBL, BindingDB, and related filtered assay sources. OFoldX does not redistribute the training set.
118
 
119
  ### Training Procedure
120
 
121
+ Upstream Boltz-2 training is staged across structure, confidence, and affinity objectives. This OFoldX artifact converts a released checkpoint into `model.safetensors` plus an OFoldX manifest; it does not run Boltz-2 training.
122
 
123
  ## Evaluation
124