@nahisaho/satori 0.15.0 → 0.17.0

package/src/.github/skills/scientific-parasite-genomics/SKILL.md

@@ -0,0 +1,280 @@
+---
+name: scientific-parasite-genomics
+description: |
+  寄生虫ゲノミクススキル。PlasmoDB/VectorBase/ToxoDB REST API
+  による寄生虫ゲノム検索・遺伝子情報・薬剤標的同定・比較
+  ゲノミクス。直接 REST API 連携 (TU 外)。
+tu_tools: []
+---
+
+# Scientific Parasite Genomics
+
+VEuPathDB ファミリー (PlasmoDB, VectorBase, ToxoDB, TriTrypDB)
+の REST API を活用した寄生虫ゲノミクス解析パイプラインを提供
+する。
+
+## When to Use
+
+- マラリア原虫ゲノム (PlasmoDB) を検索するとき
+- 蚊・ダニ等の媒介生物ゲノム (VectorBase) を検索するとき
+- トキソプラズマゲノム (ToxoDB) を検索するとき
+- トリパノソーマ/リーシュマニアゲノム (TriTrypDB) を検索するとき
+- 寄生虫の薬剤標的候補を同定するとき
+- 寄生虫間の比較ゲノミクスを実施するとき
+
+---
+
+## Quick Start
+
+## 1. VEuPathDB 遺伝子検索
+
+```python
+import requests
+import pandas as pd
+import numpy as np
+
+VEUPATHDB_SITES = {
+    "plasmo": "https://plasmodb.org/plasmo/service",
+    "vector": "https://vectorbase.org/vectorbase/service",
+    "toxo": "https://toxodb.org/toxo/service",
+    "tritryp": "https://tritrypdb.org/tritrypdb/service",
+}
+
+
+def veupathdb_search_genes(organism, query, db="plasmo",
+                             limit=100):
+    """
+    VEuPathDB — 遺伝子検索。
+
+    Parameters:
+        organism: str — 生物種名 (例: "Plasmodium falciparum 3D7")
+        query: str — 検索キーワード (例: "kinase", "transporter")
+        db: str — データベース ("plasmo", "vector", "toxo", "tritryp")
+        limit: int — 最大結果数
+    """
+    base = VEUPATHDB_SITES.get(db, VEUPATHDB_SITES["plasmo"])
+    url = f"{base}/record-types/gene/searches/GenesByTextSearch"
+
+    payload = {
+        "searchConfig": {
+            "parameters": {
+                "text_expression": query,
+                "text_fields": "Gene ID,Gene Name or Symbol,"
+                                "Gene product",
+                "organism": [organism],
+            }
+        },
+        "reportConfig": {
+            "attributes": ["primary_key", "gene_name",
+                           "gene_product", "gene_type",
+                           "chromosome", "start_min",
+                           "end_max", "strand"],
+            "pagination": {"offset": 0, "numRecords": limit},
+        },
+    }
+    headers = {"Content-Type": "application/json"}
+    resp = requests.post(url, json=payload, headers=headers,
+                          timeout=60)
+    resp.raise_for_status()
+    data = resp.json()
+
+    results = []
+    for rec in data.get("records", []):
+        attrs = rec.get("attributes", {})
+        results.append({
+            "gene_id": attrs.get("primary_key", ""),
+            "gene_name": attrs.get("gene_name", ""),
+            "product": attrs.get("gene_product", ""),
+            "gene_type": attrs.get("gene_type", ""),
+            "chromosome": attrs.get("chromosome", ""),
+            "start": attrs.get("start_min", None),
+            "end": attrs.get("end_max", None),
+            "strand": attrs.get("strand", ""),
+        })
+
+    df = pd.DataFrame(results)
+    print(f"VEuPathDB ({db}) genes: {len(df)} results "
+          f"(organism={organism}, query={query})")
+    return df
+```
+
+## 2. 遺伝子機能アノテーション
+
+```python
+def veupathdb_gene_annotation(gene_id, db="plasmo"):
+    """
+    VEuPathDB — 遺伝子機能アノテーション取得。
+
+    Parameters:
+        gene_id: str — 遺伝子 ID (例: "PF3D7_1133400")
+        db: str — データベース
+    """
+    base = VEUPATHDB_SITES.get(db, VEUPATHDB_SITES["plasmo"])
+    url = f"{base}/record-types/gene/records/{gene_id}"
+
+    params = {
+        "attributes": "all",
+        "tables": "GoTerms,InterPro,MetabolicPathways,"
+                  "PubMed,EcNumber",
+    }
+    resp = requests.get(url, params=params, timeout=30)
+    resp.raise_for_status()
+    data = resp.json()
+
+    attrs = data.get("attributes", {})
+    tables = data.get("tables", {})
+
+    annotation = {
+        "gene_id": gene_id,
+        "gene_name": attrs.get("gene_name", ""),
+        "product": attrs.get("gene_product", ""),
+        "molecular_weight": attrs.get("molecular_weight", ""),
+        "isoelectric_point": attrs.get("isoelectric_point", ""),
+        "signal_peptide": attrs.get("signal_peptide", ""),
+        "transmembrane_domains": attrs.get("transmembrane_domains", ""),
+    }
+
+    # GO Term 取得
+    go_terms = []
+    for go_rec in tables.get("GoTerms", []):
+        go_terms.append({
+            "go_id": go_rec.get("go_id", ""),
+            "go_term": go_rec.get("go_term_name", ""),
+            "ontology": go_rec.get("ontology", ""),
+            "evidence": go_rec.get("evidence_code", ""),
+        })
+    annotation["go_terms"] = go_terms
+
+    # InterPro ドメイン
+    domains = []
+    for d in tables.get("InterPro", []):
+        domains.append({
+            "interpro_id": d.get("interpro_primary_id", ""),
+            "name": d.get("interpro_name", ""),
+            "description": d.get("interpro_description", ""),
+        })
+    annotation["domains"] = domains
+
+    print(f"VEuPathDB annotation: {gene_id}, "
+          f"{len(go_terms)} GO terms, {len(domains)} domains")
+    return annotation
+```
+
+## 3. 薬剤標的候補スクリーニング
+
+```python
+def parasite_drug_target_screen(organism, db="plasmo",
+                                  essentiality_threshold=0.5):
+    """
+    寄生虫ゲノム — 薬剤標的候補スクリーニング。
+
+    Parameters:
+        organism: str — 生物種
+        db: str — データベース
+        essentiality_threshold: float — 必須性スコア閾値
+    """
+    # キナーゼ検索
+    kinases = veupathdb_search_genes(organism, "kinase", db=db)
+    # プロテアーゼ検索
+    proteases = veupathdb_search_genes(organism, "protease", db=db)
+    # トランスポーター検索
+    transporters = veupathdb_search_genes(
+        organism, "transporter", db=db)
+
+    all_targets = pd.concat([kinases, proteases, transporters],
+                              ignore_index=True)
+    all_targets = all_targets.drop_duplicates(subset=["gene_id"])
+
+    # 薬剤標的性スコア (ヒューリスティック)
+    all_targets["target_class"] = "unknown"
+    all_targets.loc[
+        all_targets["gene_id"].isin(kinases["gene_id"]),
+        "target_class"] = "kinase"
+    all_targets.loc[
+        all_targets["gene_id"].isin(proteases["gene_id"]),
+        "target_class"] = "protease"
+    all_targets.loc[
+        all_targets["gene_id"].isin(transporters["gene_id"]),
+        "target_class"] = "transporter"
+
+    print(f"Drug target screen: {len(all_targets)} candidates "
+          f"(kinases={len(kinases)}, proteases={len(proteases)}, "
+          f"transporters={len(transporters)})")
+    return all_targets
+```
+
+## 4. 寄生虫ゲノミクス統合パイプライン
+
+```python
+def parasite_genomics_pipeline(organism, query,
+                                  db="plasmo",
+                                  output_dir="results"):
+    """
+    寄生虫ゲノミクス統合パイプライン。
+
+    Parameters:
+        organism: str — 生物種 (例: "Plasmodium falciparum 3D7")
+        query: str — 検索クエリ
+        db: str — データベース
+        output_dir: str — 出力ディレクトリ
+    """
+    from pathlib import Path
+    output_dir = Path(output_dir)
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+    # 1) 遺伝子検索
+    genes = veupathdb_search_genes(organism, query, db=db)
+    genes.to_csv(output_dir / "genes.csv", index=False)
+
+    # 2) トップ遺伝子のアノテーション
+    annotations = []
+    for gene_id in genes["gene_id"].head(10):
+        try:
+            ann = veupathdb_gene_annotation(gene_id, db=db)
+            annotations.append(ann)
+        except Exception:
+            continue
+    ann_df = pd.DataFrame([{
+        k: v for k, v in a.items()
+        if not isinstance(v, list)
+    } for a in annotations])
+    ann_df.to_csv(output_dir / "annotations.csv", index=False)
+
+    # 3) 薬剤標的スクリーニング
+    targets = parasite_drug_target_screen(organism, db=db)
+    targets.to_csv(output_dir / "drug_targets.csv", index=False)
+
+    print(f"Parasite genomics pipeline: {output_dir}")
+    return {
+        "genes": genes,
+        "annotations": annotations,
+        "drug_targets": targets,
+    }
+```
+
+---
+
+## ToolUniverse 連携
+
+直接 REST API 使用 (VEuPathDB は ToolUniverse 外)。
+
+## パイプライン統合
+
+```
+infectious-disease → parasite-genomics → phylogenetics
+  (病原体情報)      (寄生虫ゲノム)       (系統解析)
+       │                   │                  ↓
+  drug-discovery ─────────┘           comparative-genomics
+  (薬剤探索)           │              (比較ゲノミクス)
+                        ↓
+                  pathway-enrichment
+                  (パスウェイ解析)
+```
+
+## パイプライン出力
+
+| ファイル | 説明 | 次スキル |
+|---------|------|---------|
+| `results/genes.csv` | 遺伝子一覧 | → phylogenetics |
+| `results/annotations.csv` | 機能アノテーション | → pathway-enrichment |
+| `results/drug_targets.csv` | 薬剤標的候補 | → drug-discovery |

package/src/.github/skills/scientific-plant-biology/SKILL.md

@@ -0,0 +1,321 @@
+---
+name: scientific-plant-biology
+description: |
+  植物バイオロジー統合スキル。Plant Reactome 代謝パスウェイ・
+  TAIR Arabidopsis ゲノム情報・Phytozome 比較ゲノミクス・
+  Ensembl Plants 種間オーソログ解析。
+---
+
+# Scientific Plant Biology
+
+Plant Reactome / TAIR / Phytozome / Ensembl Plants を活用した
+植物ゲノム・代謝パスウェイ統合解析パイプラインを提供する。
+
+## When to Use
+
+- 植物代謝パスウェイ解析 (Plant Reactome) を実行するとき
+- Arabidopsis thaliana の遺伝子・タンパク質情報を取得するとき
+- 植物種間の比較ゲノミクス解析を行うとき
+- 植物オーソログ・パラログを同定するとき
+- 作物改良のための候補遺伝子を探索するとき
+- 植物表現型データと遺伝子型を統合するとき
+
+---
+
+## Quick Start
+
+## 1. Plant Reactome パスウェイ検索
+
+```python
+import requests
+import pandas as pd
+import json
+
+PLANT_REACTOME = "https://plantreactome.gramene.org/ContentService"
+
+
+def plant_reactome_search(query, species="Oryza sativa"):
+    """
+    Plant Reactome — 植物代謝/シグナルパスウェイ検索。
+
+    Parameters:
+        query: str — 検索クエリ (例: "photosynthesis")
+        species: str — 種名
+    """
+    url = f"{PLANT_REACTOME}/search/query"
+    params = {"query": query, "species": species, "cluster": True}
+    resp = requests.get(url, params=params, timeout=30)
+    resp.raise_for_status()
+    data = resp.json()
+
+    results = []
+    for group in data.get("results", []):
+        for entry in group.get("entries", []):
+            results.append({
+                "stId": entry.get("stId", ""),
+                "name": entry.get("name", ""),
+                "species": entry.get("species", ""),
+                "type": entry.get("exactType", ""),
+                "compartment": entry.get("compartmentNames", []),
+            })
+
+    df = pd.DataFrame(results)
+    print(f"Plant Reactome: '{query}' → {len(df)} entries ({species})")
+    return df
+
+
+def plant_reactome_pathway_detail(pathway_id):
+    """
+    Plant Reactome パスウェイ詳細取得。
+
+    Parameters:
+        pathway_id: str — パスウェイ ID (例: "R-OSA-1119616")
+    """
+    url = f"{PLANT_REACTOME}/data/pathway/{pathway_id}/containedEvents"
+    resp = requests.get(url, timeout=30)
+    resp.raise_for_status()
+    events = resp.json()
+
+    steps = []
+    for event in events:
+        steps.append({
+            "stId": event.get("stId", ""),
+            "name": event.get("displayName", ""),
+            "type": event.get("className", ""),
+            "input_count": len(event.get("input", [])),
+            "output_count": len(event.get("output", [])),
+            "catalyst": event.get("catalystActivity", [{}])[0].get(
+                "displayName", "") if event.get("catalystActivity") else "",
+        })
+
+    df = pd.DataFrame(steps)
+    print(f"Pathway {pathway_id}: {len(df)} reaction steps")
+    return df
+```
+
+## 2. TAIR Arabidopsis 遺伝子情報
+
+```python
+TAIR_BASE = "https://www.arabidopsis.org/api"
+
+
+def tair_gene_search(gene_id=None, gene_name=None, keyword=None):
+    """
+    TAIR — Arabidopsis thaliana 遺伝子情報取得。
+
+    Parameters:
+        gene_id: str — AGI ID (例: "AT1G01010")
+        gene_name: str — 遺伝子名 (例: "FLC")
+        keyword: str — キーワード検索
+    """
+    if gene_id:
+        url = f"{TAIR_BASE}/gene/{gene_id}"
+        resp = requests.get(url, timeout=30)
+        resp.raise_for_status()
+        data = resp.json()
+        return pd.DataFrame([{
+            "agi_id": data.get("locus", ""),
+            "name": data.get("name", ""),
+            "description": data.get("description", ""),
+            "chromosome": data.get("chromosome", ""),
+            "start": data.get("start", ""),
+            "end": data.get("end", ""),
+            "strand": data.get("strand", ""),
+            "gene_model_type": data.get("gene_model_type", ""),
+        }])
+
+    # キーワード検索
+    search_term = gene_name or keyword or ""
+    url = f"{TAIR_BASE}/search/gene"
+    params = {"query": search_term, "limit": 50}
+    resp = requests.get(url, params=params, timeout=30)
+    resp.raise_for_status()
+    data = resp.json()
+
+    results = []
+    for gene in data.get("results", []):
+        results.append({
+            "agi_id": gene.get("locus", ""),
+            "name": gene.get("name", ""),
+            "description": gene.get("description", ""),
+            "chromosome": gene.get("chromosome", ""),
+        })
+
+    df = pd.DataFrame(results)
+    print(f"TAIR: '{search_term}' → {len(df)} genes")
+    return df
+
+
+def tair_gene_expression(gene_id):
+    """
+    TAIR — 遺伝子発現パターン取得。
+
+    Parameters:
+        gene_id: str — AGI ID
+    """
+    url = f"{TAIR_BASE}/gene/{gene_id}/expression"
+    resp = requests.get(url, timeout=30)
+    resp.raise_for_status()
+    data = resp.json()
+
+    tissues = []
+    for expr in data.get("expression", []):
+        tissues.append({
+            "tissue": expr.get("tissue", ""),
+            "stage": expr.get("developmental_stage", ""),
+            "level": expr.get("expression_level", ""),
+            "source": expr.get("source", ""),
+        })
+
+    df = pd.DataFrame(tissues)
+    print(f"TAIR expression: {gene_id} → {len(df)} tissue records")
+    return df
+```
+
+## 3. Ensembl Plants 種間比較
+
+```python
+ENSEMBL_PLANTS = "https://rest.ensembl.org"
+
+
+def ensembl_plants_orthologs(gene_id, source_species="arabidopsis_thaliana",
+                              target_species=None):
+    """
+    Ensembl Plants — 植物種間オーソログ検索。
+
+    Parameters:
+        gene_id: str — Ensembl Gene ID or symbol
+        source_species: str — 起源種
+        target_species: str — ターゲット種 (None = 全種)
+    """
+    url = f"{ENSEMBL_PLANTS}/homology/id/{gene_id}"
+    params = {
+        "type": "orthologues",
+        "content-type": "application/json",
+        "compara": "plants",
+    }
+    if target_species:
+        params["target_species"] = target_species
+
+    resp = requests.get(url, params=params, timeout=30)
+    resp.raise_for_status()
+    data = resp.json()
+
+    orthologs = []
+    for homology in data.get("data", [{}])[0].get("homologies", []):
+        target = homology.get("target", {})
+        orthologs.append({
+            "source_gene": gene_id,
+            "source_species": source_species,
+            "target_gene": target.get("id", ""),
+            "target_species": target.get("species", ""),
+            "target_protein": target.get("protein_id", ""),
+            "identity": target.get("perc_id", 0),
+            "dn_ds": homology.get("dn_ds", None),
+            "type": homology.get("type", ""),
+        })
+
+    df = pd.DataFrame(orthologs)
+    print(f"Ensembl Plants orthologs: {gene_id} → {len(df)} homologs")
+    return df
+```
+
+## 4. Phytozome 比較ゲノミクス
+
+```python
+PHYTOZOME_BASE = "https://phytozome-next.jgi.doe.gov/api"
+
+
+def phytozome_gene_family(gene_id, species="Athaliana"):
+    """
+    Phytozome — 遺伝子ファミリー・比較ゲノミクス。
+
+    Parameters:
+        gene_id: str — 遺伝子 ID
+        species: str — 種略称
+    """
+    url = f"{PHYTOZOME_BASE}/search"
+    params = {"query": gene_id, "organism": species}
+    resp = requests.get(url, params=params, timeout=30)
+    resp.raise_for_status()
+    data = resp.json()
+
+    families = []
+    for hit in data.get("hits", []):
+        families.append({
+            "gene_id": hit.get("gene_id", ""),
+            "family_id": hit.get("family_id", ""),
+            "family_name": hit.get("family_name", ""),
+            "species": hit.get("organism", ""),
+            "annotation": hit.get("annotation", ""),
+            "pfam_domains": hit.get("pfam", []),
+        })
+
+    df = pd.DataFrame(families)
+    print(f"Phytozome: {gene_id} → {len(df)} family members")
+    return df
+```
+
+## 5. 植物バイオロジー統合パイプライン
+
+```python
+def plant_biology_pipeline(gene_query, species="Oryza sativa",
+                            output_dir="results"):
+    """
+    植物バイオロジー統合パイプライン。
+
+    Parameters:
+        gene_query: str — 遺伝子/パスウェイクエリ
+        species: str — 対象種
+        output_dir: str — 出力ディレクトリ
+    """
+    from pathlib import Path
+    output_dir = Path(output_dir)
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+    # 1) Plant Reactome パスウェイ
+    pathways = plant_reactome_search(gene_query, species=species)
+    pathways.to_csv(output_dir / "plant_pathways.csv", index=False)
+
+    # 2) TAIR (Arabidopsis ならば)
+    tair_genes = tair_gene_search(keyword=gene_query)
+    tair_genes.to_csv(output_dir / "tair_genes.csv", index=False)
+
+    # 3) Ensembl Plants オーソログ
+    if len(tair_genes) > 0:
+        top_gene = tair_genes.iloc[0]["agi_id"]
+        orthologs = ensembl_plants_orthologs(top_gene)
+        orthologs.to_csv(output_dir / "orthologs.csv", index=False)
+    else:
+        orthologs = pd.DataFrame()
+
+    print(f"Plant biology pipeline: {output_dir}")
+    return {
+        "pathways": pathways,
+        "tair_genes": tair_genes,
+        "orthologs": orthologs,
+    }
+```
+
+---
+
+## パイプライン統合
+
+```
+pathway-enrichment → plant-biology → environmental-ecology
+  (KEGG/Reactome)   (PlantReactome)   (生態学/環境)
+       │                   │                ↓
+  gene-annotation ────────┘         marine-ecology
+  (GO/InterPro)      │              (OBIS/WoRMS)
+                      ↓
+               comparative-genomics
+               (Ensembl 比較)
+```
+
+## パイプライン出力
+
+| ファイル | 説明 | 次スキル |
+|---------|------|---------|
+| `results/plant_pathways.csv` | Plant Reactome パスウェイ | → pathway-enrichment |
+| `results/tair_genes.csv` | TAIR Arabidopsis 遺伝子 | → gene-annotation |
+| `results/orthologs.csv` | 種間オーソログ | → comparative-genomics |