Datasets:

Dataset Viewer
The dataset could not be loaded because the splits use different data file formats, which is not supported. Read more about the splits configuration. Click for more details.
Couldn't infer the same data file format for all splits. Got {NamedSplit('validation'): ('json', {}), NamedSplit('test'): ('text', {})}
Error code:   FileFormatMismatchBetweenSplitsError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/datasets-cards)

Files

辅助文件:

  • pistachio_rxn_name_def_map.json
    • "2.1.2": "Carboxylic acid + amine condensation"

重要文件:

  • coley_rxn_type_lib.json

    • "REACTION_TYPE": List[
          "A>B>C [optional |f:x.x|]" 
      ]
      # A: reactants
      # B: reagents
      # C: product
      
  • coley_rxn_type_lib_wo_agent_smiles.json

    • w/o reagents, only "A>>C"
  • Coley_reaction_example.json

    • ID[int]: {
          "name_rxn": "2.1.2",
          "name_rxn_def": "Carboxylic acid + amine condensation",
          "smiles": A>B>C
      }
      
    • Sub-set of pistachio_rxn_type_lib.json (only included in coley paper)

  • mech_total.json

    • "REACTION_TYPE": List[
          "pickle_path":  "/cto_labs/AIDD/DATA/React/Mechanism/pistachio/single/CO2H-tBu deprotection/type_1/513.pickle",
          "atom_map_rxn": {
              "reactants": List[SMILES-AtomMap],
              "products": List[SMILES-AtomMap],
              "byproducts": List[SMILES-AtomMap],
          },
          "unmapped_rxn": {
              "reactants": List[SMILES],
              "products": List[SMILES],
              "byproducts": List[SMILES],
          }
          ...
      ]
      ...
      
      • mech_example.json: each rxn-type contains 1 example

重要Dir

  • single/

    each folder is a leaf-type reaction

    • # 89 types
      - Alcohol + amine condensation/
          - type_0
              - xxx.pickle # mechanism step graph
              - xxx.txt # mechanism steps
          - ...
      ...
      
  • test5/ (similar to test50/)

    random sample 5 instances for each leaf-type of reaction from coley_rxn_type_lib.json

    • taxonomy/

      # 5 major rxn types (define by Bing Feng + Claude)
      Carbon-Carbon Bond Formation/
          - unmapped_rxn.txt
      Carbon-Heteroatom Bond Formation/
          - unmapped_rxn.txt
      Deprotection Reactions/
          - unmapped_rxn.txt
      Functional Group Interconversion/
          - unmapped_rxn.txt
      Other Reactions/
          - unmapped_rxn.txt
      
    • template/

      - indices.json # each rxn type extracted 5 ids, correspond to id in coley_rxn_type_lib.json. To keep traceable.
      - templates.json
          "REACTION_TYPE": List[
              {
                  "reactants_smiles_am": "CC(C)(C)[O:1][CH:2]=[O:3]", # reactants smiles based "reactants_smarts", maybe not valid, w/ atomMap
                  "products_smiles_am": "[O:1][CH:2]=[O:3]", # similar to "reactants_smiles_am"
                  "reactants_smiles": "CC(C)(C)OC=O", # reactants_smiles_am w/o atomMap
                  "product_smiles": "OC=O", # products_smiles_am w/o atomAmp
                  "reactants_smarts": "C-C(-C)(-C)-[O;H0;D2;+0:1]-[C:2]=[O;D1;H0:3]", # extracted using coley rdChiral template extracion function
                  "products_smarts": "[O;D1;H0:3]=[C:2]-[OH;D1;+0:1]",
                  "reagents_smiles_am": "COc1ccccc1.O=C(O)C(F)(F)F", # w/ atomMap
                  "rxn_am": "[CH2:1]=[CH:2][C@:3]1([S:4][C@@H:5]2[C@@H:6]([C:7]([N:8]2[CH:9]1[C:10](=[O:11])[O:12]C(C)(C)C)=[O:13])[NH:14][C:15](=[O:16])[C@H:17]([NH2:18])[c:19]1[cH:20][cH:21][cH:22][cH:23][cH:24]1)[CH3:25]>>[CH2:1]=[CH:2][C@:3]1([S:4][C@@H:5]2[C@@H:6]([C:7]([N:8]2[CH:9]1[C:10](=[O:11])[OH:12])=[O:13])[NH:14][C:15](=[O:16])[C@H:17]([NH2:18])[c:19]1[cH:20][cH:21][cH:22][cH:23][cH:24]1)[CH3:25]", # rxn w/ atomMap
                  "rxn_unmapped": "C=C[C@@]1(C)S[C@@H]2[C@H](NC(=O)[C@H](N)c3ccccc3)C(=O)N2C1C(=O)OC(C)(C)C>>C=C[C@@]1(C)S[C@@H]2[C@H](NC(=O)[C@H](N)c3ccccc3)C(=O)N2C1C(=O)O" # w/o atomMap
              },
              ...
          ]
      
  • taxonomy/

    • 五大类每一类的train.json and test.json

      Major Types # Types # Train # Test
      Carbon-Carbon Bond Formation 25 94,543 1250
      Carbon-Heteroatom Bond Formation 28 567,574 1400
      Deprotection Reactions 10 243,266 500
      Functional Group Interconversion 19 189,296 950
      Other Reactions 3 21,917 150

      Total # Types = 85 Total # Train = 1,116,596 Total # Test = 4,250

    • json List[Dict], 信息来源mech_total.json

Downloads last month
1

Collection including IDEA-AI4S/rxn_mech