AGENT SKILL

molclaw-mol-similarity

Calculate both Tanimoto similarities and the count of shared structural fragments between a target molecule and a list of candidate molecules via Morgan fingerprints.

License 许可
MIT
最近更新
2026-10-09
Stars
★ 0
Forks
0

01 前言 Front Matter

name
molclaw-mol-similarity
license
MIT license
metadata
{ "skill-author": "PJLab" }
description
Calculate both Tanimoto similarities and the count of shared structural fragments between a target molecule and a list of candidate molecules via Morgan fingerprints.

03 SKILL.md 正文

Molecule Similarity Calculation

Note:

  • Local files are not directly accessible by the server. Please upload them to the server using molclaw-file-transfer before execution.
  • For PDB file inputs, it is recommended to preprocess them using molclaw-pdbfixer before execution.
  • Please refer to skill molclaw-scp-server to complete tool invocation.

Scene 1: Compute the Tanimoto similarities between a target molecule and a list of candidate molecules using Morgan fingerprints. Need to use the tool calculate_morgan_fingerprint_similarity.

The description of tool calculate_morgan_fingerprint_similarity.

Compute the Tanimoto similarities between a target molecule and a list of candidate molecules using Morgan fingerprints.
Args:
    target_smiles (str): SMILES string of the target molecule
    candidate_smiles_list (List[str]): List of candidate molecule SMILES strings
    radius (int): Morgan fingerprint radius, default is 2
    nBits (int): Morgan fingerprint vector bits number, default is 2048
Return:
    status (str): success/error
    msg (str): message
    similarities (List[dict]): List of dict, each containing the keys 'smiles' and 'score'.
        --smiles (str): A SMILES string of candidate_smiles_list
        --score (float): Similarity value between the candidate SMILES and the target SMILES

How to use tool calculate_morgan_fingerprint_similarity :

response = await client.session.call_tool(
    "calculate_morgan_fingerprint_similarity",
    arguments={
        "target_smiles": target_smiles,
        "candidate_smiles_list": candidate_smiles_list,
        "radius": radius,
        "nBits": nBits
    }
)
result = client.parse_result(response)
similarities = result["similarities"]

Scene 2: Compute the count of shared structural fragments between a target molecule and a list of candidate molecules using Morgan fingerprints. Need to use the tool calculate_common_fragments.

The description of tool calculate_common_fragments.

Compute the count of shared structural fragments between a target molecule and a list of candidate molecules using Morgan fingerprints.
Args:
    target_smiles (str): SMILES string of the target molecule
    candidate_smiles_list (List[str]): List of candidate molecule SMILES strings
    radius (int): Morgan fingerprint radius, default is 2
Return:
    status (str): success/error
    msg (str): message
    fragments_info (List[dict]): List of dict, each containing the keys 'smiles' and 'common_fragment_count'.
        --smiles (str): A SMILES string of candidate_smiles_list
        --common_fragment_count (float): Number of structural fragments shared between the candidate SMILES and the target SMILES

How to use tool calculate_common_fragments :

response = await client.session.call_tool(
    "calculate_common_fragments",
    arguments={
        "target_smiles": target_smiles,
        "candidate_smiles_list": candidate_smiles_list,
        "radius": radius
    }
)
result = client.parse_result(response)
fragments_info = result["fragments_info"]

04 同仓库更多 Skills