Molclaw Peptide Sampling

Peptide Molecule Generation

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.

The description of tool pepinventpeptidesamplingbypeptide.

Generate new peptide molecules sampling from the input peptide sequence.
Args:
    peptide (str): SMILES representation of a peptide sequence, with amino acid residues separated by '|?|', e.g., 'N[C@@H](CCCCN)C(=O)|?|N[C@@H](CC(C)C)C(=O)|?|N[C@@H](CCCNC(=N)N)C(=O)' 
    n (int): Number of molecules for sampling
    filter_preset (str): Filter preset, options: ['none', 'minimal', 'default', 'strict'], default is 'default'
    mw_min (float): Minimum molecular weight, default is 0.0
    mw_max (float): Maximum molecular weight, default is 0.0
Return:
    status (str): success/error
    msg (str): message
    save_smiles_file (str): Path to the saved SMILES file
    output_smiles_list (List[str]): List of generated SMILES strings

How to use tool pepinventpeptidesamplingbypeptide :

response = await client.session.call_tool(
    "pepinvent_peptide_sampling_by_peptide",
    arguments={
        "peptide": smiles,
        "n": n,
        "filter_preset": filter_type,
        "mw_min": mw_min,
        "mw_max": mw_max
    }
)
result = client.parse_result(response)
output_smiles_list = result["output_smiles_list"]

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: InternScience
• Source: InternScience/MolClaw
• License: MIT

Install and usage instructions live in the source repository linked above.