メニュー
Use when running AlphaFold2 predictions on custom protein sequences, validating designed sequences via self-consistency, predicting binder-target complexes, or interpreting AF2 confidence metrics (pLDDT, pTM, ipTM).
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/lamm-mit/scienceclaw --skill alphafoldコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
SOC 職業分類に基づく
SKILL.md を表示中
| name | alphafold |
| description | Use when running AlphaFold2 predictions on custom protein sequences, validating designed sequences via self-consistency, predicting binder-target complexes, or interpreting AF2 confidence metrics (pLDDT, pTM, ipTM). |
| metadata | null |
Use when the agent needs to run AlphaFold2 for protein structure prediction and complex modeling. Covers validating designed sequences, predicting binder-target complexes, and calculating confidence metrics (pLDDT, pTM, ipTM).
Distinct from alphafold-database (which retrieves pre-computed structures) — this skill covers running AF2 predictions on custom sequences.
pip install colabfold[alphafold]
# Single chain
colabfold_batch input.fasta output_dir/ \
--model-type alphafold2_ptm \
--num-recycles 3
# Complex (multimer) — comma-separate chains in FASTA header
# >complex:ChainA,ChainB
colabfold_batch complex.fasta output_dir/ \
--model-type alphafold2_multimer_v3 \
--num-recycles 20 \
--num-models 5
# Install
wget https://raw.githubusercontent.com/YoshitakaMo/localcolabfold/main/install_colabbatch_linux.sh
bash install_colabbatch_linux.sh
# Run offline
colabfold_batch sequences.fasta results/ \
--model-type alphafold2_multimer_v3 \
--num-recycles 3 \
--use-gpu-relax
pip install openfold
python run_pretrained_openfold.py \
--fasta_paths input.fasta \
--output_dir results/ \
--model_device cuda:0
| Parameter | Values | Notes |
|---|---|---|
--model-type | alphafold2_ptm, alphafold2_multimer_v3 | Use multimer for complexes |
--num-recycles | 3–20 | More recycles = better accuracy, slower |
--num-models | 1–5 | 5 models for ensemble confidence |
--msa-mode | mmseqs2_uniref_env (default), single_sequence | Single = no MSA, faster |
--use-gpu-relax | flag | Amber relaxation on GPU |
import numpy as np
import json
# Load result JSON
with open("result_model_1.json") as f:
result = json.load(f)
plddt = np.array(result["plddt"]) # Per-residue confidence 0-100
ptm = result["ptm"] # Global TM-score estimate 0-1
iptm = result.get("iptm", None) # Interface TM-score (multimer only)
pae = np.array(result.get("pae", [])) # Predicted Aligned Error matrix
# Quality thresholds
print(f"Mean pLDDT: {plddt.mean():.1f}") # >70 = good, >90 = excellent
print(f"pTM: {ptm:.3f}") # >0.5 = confident fold
if iptm:
print(f"ipTM: {iptm:.3f}") # >0.6 = reliable complex, >0.8 = high confidence
# Design → predict → measure similarity to input backbone
# 1. Generate sequences with ProteinMPNN
# 2. Predict structure of each sequence with AF2
# 3. Calculate TM-score / RMSD vs. design backbone
python3 -c "
from Bio.PDB import PDBParser, Superimposer
# Compare predicted vs. designed structure
# High TM-score (>0.8) = sequence encodes target fold
"
| File | Contents |
|---|---|
*_relaxed_rank_1.pdb | Top-ranked relaxed structure |
*_unrelaxed_rank_1.pdb | Top-ranked unrelaxed structure |
result_model_*.json | Scores: pLDDT, pTM, ipTM, PAE matrix |
*_coverage.png | MSA coverage plot |
*_pae.png | PAE heatmap (low = confident) |
| Metric | Poor | Acceptable | Good | Excellent |
|---|---|---|---|---|
| Mean pLDDT | <50 | 50–70 | 70–90 | >90 |
| pTM | <0.4 | 0.4–0.5 | 0.5–0.7 | >0.7 |
| ipTM (complex) | <0.5 | 0.5–0.6 | 0.6–0.8 | >0.8 |
| Interface PAE | >20 Å | 15–20 Å | 8–15 Å | <8 Å |
| Problem | Cause | Fix |
|---|---|---|
| Low ipTM despite high pLDDT | Chains fold well independently but don't interact | Redesign interface residues |
| High PAE at interface | Interface not well-determined | Add more recycles; check contact predictions |
| OOM on GPU | Sequence too long | Use --chunk-size 128 or CPU for MSA |
| All models disagree | Disordered region or wrong fold | Check MSA depth; try --msa-mode single_sequence |