"""Survey: are well-known protein ligands annotated involved_in the pathway they activate? For each ligand (and a few receptors as comparators) in human, Drosophila and S. pombe, fetch all biological-process annotations from QuickGO and classify each against the ligand's own pathway term ("cognate" pathway): pathway the term is the cognate pathway or an is_a/part_of descendant regulation the term reaches the cognate pathway only through regulates relations other any other signal transduction term Accessions are resolved from UniProt (reviewed entries; a few fly genes need the unreviewed FlyBase-mapped accession, given in OVERRIDES). The ligand and pathway lists are a hand-picked sample of textbook pathways. Usage: uv run python projects/LIGANDS_IN_SIGNALING/scripts/survey/ligand_pathway_survey.py \ --out projects/LIGANDS_IN_SIGNALING/results/ligand_survey.json """ from __future__ import annotations import argparse import collections import json from pathlib import Path import requests QUICKGO = "https://www.ebi.ac.uk/QuickGO/services" HEADERS = {"Accept": "application/json"} # gene -> cognate pathway term COGNATE = { "9606": { "WNT3A": "GO:0016055", "WNT5A": "GO:0016055", "WNT1": "GO:0016055", "SHH": "GO:0007224", "IHH": "GO:0007224", "DLL1": "GO:0007219", "JAG1": "GO:0007219", "TGFB1": "GO:0007179", "BMP4": "GO:0030509", "NODAL": "GO:0038092", "EGF": "GO:0007173", "TGFA": "GO:0007173", "INS": "GO:0008286", "FGF2": "GO:0008543", "FGF8": "GO:0008543", "IFNG": "GO:0060333", "TNF": "GO:0033209", "IL6": "GO:0070102", "KITLG": "GO:0038109", "EPO": "GO:0038162", "CXCL12": "GO:0038146", "VEGFA": "GO:0048010", "NGF": "GO:0038180", "LEP": "GO:0033210", # receptor comparators "FZD1": "GO:0016055", "PTCH1": "GO:0007224", "NOTCH1": "GO:0007219", "EGFR": "GO:0007173", "INSR": "GO:0008286", }, "7227": { "wg": "GO:0016055", "hh": "GO:0007224", "Dl": "GO:0007219", "Ser": "GO:0007219", "dpp": "GO:0030509", "gbb": "GO:0030509", "spi": "GO:0007173", "grk": "GO:0007173", "upd1": "GO:0007259", "Ilp2": "GO:0008286", "boss": "GO:0045500", "Pvf1": "GO:0048010", "bnl": "GO:0008543", "spz": "GO:0008063", "fz": "GO:0016055", "ptc": "GO:0007224", "N": "GO:0007219", "Egfr": "GO:0007173", "InR": "GO:0008286", "Tl": "GO:0008063", }, # The pombe pheromone pathway term (GO:0000750) is obsolete; no cognate term is checked. "284812": {"map2": None, "mfm1": None, "mfm2": None, "mfm3": None, "mam2": None, "map3": None}, } RECEPTORS = {"FZD1", "PTCH1", "NOTCH1", "EGFR", "INSR", "fz", "ptc", "N", "Egfr", "InR", "Tl", "mam2", "map3"} # Name search returns the wrong gene (Dl matches dorsal) or no reviewed entry. OVERRIDES = {("7227", "Dl"): "P10041", ("7227", "upd1"): "Q9VWW4", ("7227", "Pvf1"): "Q9VWP6", ("7227", "bnl"): "Q9VDT9"} SIGNAL_TRANSDUCTION = "GO:0007165" def resolve(taxon: str, gene: str) -> str | None: if (taxon, gene) in OVERRIDES: return OVERRIDES[(taxon, gene)] resp = requests.get("https://rest.uniprot.org/uniprotkb/search", params={ "query": f"gene_exact:{gene} AND organism_id:{taxon} AND reviewed:true", "fields": "accession,gene_names", "format": "json"}, timeout=30) resp.raise_for_status() r = resp.json() hits = [(x["primaryAccession"], (x.get("genes") or [{}])[0].get("geneName", {}).get("value")) for x in r.get("results", [])] exact = [a for a, n in hits if n == gene] return (exact or [a for a, _ in hits] or [None])[0] def bp_annotations(acc: str) -> list[dict]: rows, page = [], 1 while True: resp = requests.get(f"{QUICKGO}/annotation/search", headers=HEADERS, params={ "geneProductId": acc, "aspect": "biological_process", "limit": 200, "page": page}, timeout=30) resp.raise_for_status() r = resp.json() rows += r["results"] if page >= r["pageInfo"]["total"]: return rows page += 1 def ancestors(terms: list[str], relations: str) -> dict[str, set[str]]: out = {} for i in range(0, len(terms), 50): chunk = ",".join(terms[i:i + 50]) resp = requests.get(f"{QUICKGO}/ontology/go/terms/{chunk}/ancestors", headers=HEADERS, params={"relations": relations}, timeout=30) resp.raise_for_status() r = resp.json() for x in r["results"]: out[x["id"]] = set(x.get("ancestors", [])) | {x["id"]} return out def main() -> None: ap = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter) ap.add_argument("--out", type=Path, required=True) args = ap.parse_args() annots = {} for taxon, genes in COGNATE.items(): for gene in genes: acc = resolve(taxon, gene) annots[(taxon, gene)] = (acc, bp_annotations(acc) if acc else []) terms = sorted({a["goId"] for _, rows in annots.values() for a in rows}) structural = ancestors(terms, "is_a,part_of") with_reg = ancestors(terms, "is_a,part_of,regulates") exp = {"IDA", "IMP", "IPI", "IGI", "IEP", "EXP"} results = [] for (taxon, gene), (acc, rows) in annots.items(): cognate = COGNATE[taxon][gene] counts = collections.Counter() evidence = collections.Counter() for a in rows: t = a["goId"] if cognate and cognate in structural.get(t, ()): counts["pathway_" + a["qualifier"]] += 1 if a["qualifier"] == "involved_in": evidence["experimental" if a["goEvidence"] in exp else "other"] += 1 elif cognate and cognate in with_reg.get(t, ()): counts["regulation"] += 1 elif SIGNAL_TRANSDUCTION in structural.get(t, ()): counts["other_signal_transduction"] += 1 results.append(dict(taxon=taxon, gene=gene, accession=acc, role="receptor" if gene in RECEPTORS else "ligand", cognate_pathway=cognate, n_bp_annotations=len(rows), counts=dict(counts), pathway_involved_in_evidence=dict(evidence))) print(f"{taxon:7} {gene:7} {acc} pathway(involved_in)={counts['pathway_involved_in']:3} " f"regulation={counts['regulation']:2} other_signaling={counts['other_signal_transduction']:2}") args.out.parent.mkdir(parents=True, exist_ok=True) args.out.write_text(json.dumps(results, indent=1)) if __name__ == "__main__": main()