How do I install the Protein Property Prediction skill?

Run npx skillmds@latest add ai4protein/protein-property-prediction in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Protein Property Prediction safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Protein Property Prediction?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Protein Property Prediction free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under Apache-2.

Who published Protein Property Prediction?

ai4protein (@ai4protein) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Protein Property Prediction by ai4protein

Name: Protein Property Prediction
Author: ai4protein

Protein Property Prediction

Physicochemical properties, surface/SS features, and finetuned protein/residue function prediction. Use when the user asks for solubility, optimal temperature, activity/binding/conserved sites, RSA/SASA/secondary structure, or property tables from FASTA/PDB. Do NOT use for zero-shot mutation ranking (zero_shot_mutation_workflow) or custom model training (train tools / future finetune workflow).

ai4protein Updated Sep 17, 2026

File contents

Protein Property & Function Prediction

Overview

Orchestrates VenusFactory predict tools: FASTA physicochemical features, PDB RSA/SASA/SS, and finetuned sequence-level / residue-level function heads (Ankh / ESM2 / ProtT5 adapters under ckpt/).

Project Tools (VenusFactory2)

Tool	Args	When
calculate_physchem_from_fasta	`fasta_file`, optional `out_dir`	Length, MW, pI, gravy, etc.
calculate_rsa_from_pdb	`pdb_file`, `chain_id`	Relative solvent accessibility
calculate_sasa_from_pdb	`pdb_file`	Absolute SASA
calculate_ss_from_pdb	`pdb_file`, `chain_id`	Secondary structure
predict_protein_function	`fasta_file`, `task`, `model_name`	Solubility, Optimal Temperature, …
predict_residue_function	`fasta_file`, `task`, `model_name`	Activity / Binding / Conserved Site / Motif
predict_structure_esmfold	`sequence`	Need PDB before RSA/SASA/SS

Workflow

Properties from sequence

calculate_physchem_from_fasta
Optional predict_protein_function with an explicit task matching platform dataset names.

Structure features

Obtain PDB (protein_structure_pipeline or user upload).
Run calculate_rsa_from_pdb / calculate_sasa_from_pdb / calculate_ss_from_pdb as needed.

Residue sites

predict_residue_function with task in {Activity Site, Binding Site, Conserved Site, Motif}.
Cross-check with InterPro (interpro_domain_annotation) before proposing edits to catalytic residues.

Common mistakes

Inventing task names not present in platform constant.json / ckpt mapping.
Running RSA tools on FASTA without a structure.
Treating finetuned scores as wet-lab labels without reporting model/task.

---name: protein-property-predictiondescription: Physicochemical properties, surface/SS features, and finetuned protein/residue function prediction. Use when the user asks for solubility, optimal temperature, activity/binding/conserved sites, RSA/SASA/secondary structure, or property tables from FASTA/PDB. Do NOT use for zero-shot mutation ranking (zero_shot_mutation_workflow) or custom model training (train tools / future finetune workflow).license: Apache-2.0---# Protein Property & Function Prediction## OverviewOrchestrates VenusFactory `predict` tools: FASTA physicochemical features, PDB RSA/SASA/SS, and finetuned sequence-level / residue-level function heads (Ankh / ESM2 / ProtT5 adapters under `ckpt/`).## Project Tools (VenusFactory2)| Tool | Args | When ||------|------|------|| **calculate_physchem_from_fasta** | `fasta_file`, optional `out_dir` | Length, MW, pI, gravy, etc. || **calculate_rsa_from_pdb** | `pdb_file`, `chain_id` | Relative solvent accessibility || **calculate_sasa_from_pdb** | `pdb_file` | Absolute SASA || **calculate_ss_from_pdb** | `pdb_file`, `chain_id` | Secondary structure || **predict_protein_function** | `fasta_file`, `task`, `model_name` | Solubility, Optimal Temperature, … || **predict_residue_function** | `fasta_file`, `task`, `model_name` | Activity / Binding / Conserved Site / Motif || **predict_structure_esmfold** | `sequence` | Need PDB before RSA/SASA/SS |## Workflow### Properties from sequence1. `calculate_physchem_from_fasta`2. Optional `predict_protein_function` with an explicit `task` matching platform dataset names.### Structure features1. Obtain PDB (`protein_structure_pipeline` or user upload).2. Run `calculate_rsa_from_pdb` / `calculate_sasa_from_pdb` / `calculate_ss_from_pdb` as needed.### Residue sites1. `predict_residue_function` with task in {`Activity Site`, `Binding Site`, `Conserved Site`, `Motif`}.2. Cross-check with InterPro (`interpro_domain_annotation`) before proposing edits to catalytic residues.## Common mistakes- Inventing `task` names not present in platform `constant.json` / ckpt mapping.- Running RSA tools on FASTA without a structure.- Treating finetuned scores as wet-lab labels without reporting model/task.

Frequently asked questions