"""Count virus species with a recorded human host under current ICTV taxonomy. Reproduces the ProteinIQ analysis in content/guides/how-many-viruses-infect-humans.mdx. Inputs (download into one folder; raw files are not stored in the repository): virushostdb.tsv https://www.genome.jp/ftp/db/virushostdb/virushostdb.tsv virus_genome_type.tsv https://www.genome.jp/ftp/db/virushostdb/virus_genome_type.tsv VMR_MSL41.v1.20260729.xlsx https://ictv.global/sites/default/files/VMR/VMR_MSL41.v1.20260729.xlsx The published run used the Virus-Host DB files dated 2026-10-04 (dbrel.txt: RefSeq release 237, GenBank release 273.0) and ICTV VMR MSL41.v1 (VMR release 2026-07-29, DOI 10.5281/zenodo.21694279), all downloaded on 2026-10-23. Run (Python 3.11+, openpyxl for the VMR workbook): python analyze.py --input-dir RAW --output-dir . Method in brief. A Virus-Host DB record counts when its host taxid is 9606 (Homo sapiens). Each record's virus taxid is mapped to a species name through virus_genome_type.tsv, then to an ICTV MSL41 species by exact name, falling back to any GenBank accession shared with the VMR. Records that map to no ICTV species (NCBI "unclassified" groups, provisional names) are counted separately, as are ICTV species whose VMR host source is bacteria (phages of gut bacteria sampled from people). Each human-host ICTV species takes its strongest evidence tier across its human records: disease a human record names a disease (Virus-Host DB DISEASE or KEGG DISEASE) literature otherwise, a human record cites a PubMed ID annotation otherwise, the host comes only from database annotation (RefSeq, GenBank via NCBI Virus, UniProt) ICTV's own host-source column then screens the list: a species is counted only when its VMR host source includes "vertebrates" as a whole token ("invertebrates" does not match); species with human records but only bacterial, archaeal or invertebrate VMR hosts (gut phages and CRESS viruses sampled from people, an amoeba virus) are listed as excluded. One NCBI name is mapped by hand because ICTV split the species: NCBI "Orthoflavivirus encephalitidis" (RefSeq NC_001672, tick-borne encephalitis virus Neudoerfl) is ICTV "Orthoflavivirus neudoerflense" (VMR U27495). A species is "also recorded in another vertebrate" when any Virus-Host DB record of the same ICTV species has a non-human vertebrate host. A recorded host is a reported association, not proof of replication, disease or natural infection. """ import argparse import csv import hashlib import json import re from collections import Counter, defaultdict from pathlib import Path import openpyxl HUMAN = "9606" VMR_FILE = "VMR_MSL41.v1.20260729.xlsx" TIERS = ("disease", "literature", "annotation") ACCESSION = re.compile(r"\b([A-Z]{1,2}_?\d{5,8})(?:\.\d+)?\b") GENOME_GROUPS = { "dsDNA": "dsDNA", "dsDNA-RT": "Reverse-transcribing", "ssRNA-RT": "Reverse-transcribing", "dsRNA": "dsRNA", "ssRNA(+)": "ssRNA(+)", "ssRNA(-)": "ssRNA(-) and ambisense", "ssRNA(+/-)": "ssRNA(-) and ambisense", "ssRNA(-); ssRNA(+/-)": "ssRNA(-) and ambisense", } MANUAL_MAP = {"orthoflavivirus encephalitidis": "Orthoflavivirus neudoerflense"} def vmr_has_vertebrate_host(hosts): tokens = {token.strip().replace(" (S)", "") for host in hosts for token in host.split(",")} return "vertebrates" in tokens def genome_group(genome): groups = { GENOME_GROUPS.get(token) or ("ssDNA" if token.startswith("ssDNA") else "Other") for token in (part.strip() for part in genome.split(";")) } return groups.pop() if len(groups) == 1 else "Other" def read_tsv(path): with open(path, encoding="utf-8", newline="") as stream: return list(csv.DictReader(stream, delimiter="\t")) def read_vmr(path): sheet = openpyxl.load_workbook(path, read_only=True)["VMR MSL41"] rows = sheet.iter_rows(values_only=True) header = [str(cell).strip() if cell else "" for cell in next(rows)] column = {name: header.index(name) for name in ("Realm", "Family", "Species", "Genome", "Host source", "Virus GENBANK accession")} species, by_accession = {}, {} for row in rows: name = row[column["Species"]] if not name: continue name = name.strip() entry = species.setdefault(name, {"realm": row[column["Realm"]] or "", "family": row[column["Family"]] or "", "genome": set(), "host": set()}) entry["genome"].add(row[column["Genome"]] or "") entry["host"].add(row[column["Host source"]] or "") for accession in ACCESSION.findall(row[column["Virus GENBANK accession"]] or ""): by_accession[accession] = name return species, by_accession def sha256(path): return hashlib.sha256(path.read_bytes()).hexdigest() def main(): parser = argparse.ArgumentParser(description=__doc__.splitlines()[0]) parser.add_argument("--input-dir", type=Path, required=True) parser.add_argument("--output-dir", type=Path, required=True) args = parser.parse_args() records = read_tsv(args.input_dir / "virushostdb.tsv") ncbi_species = {row["virus_tax_id"]: row["species"] for row in read_tsv(args.input_dir / "virus_genome_type.tsv")} vmr, vmr_by_accession = read_vmr(args.input_dir / VMR_FILE) vmr_lower = {name.lower(): name for name in vmr} def ictv_species(record): ncbi_name = ncbi_species.get(record["virus tax id"], "").lower() name = vmr_lower.get(ncbi_name) or MANUAL_MAP.get(ncbi_name) if name: return name, "name" for accession in ACCESSION.findall(record["refseq id"]): if accession in vmr_by_accession: return vmr_by_accession[accession], "accession" return None, None human_records = [row for row in records if row["host tax id"] == HUMAN] tier = {} match_method = Counter() human_virus_taxa = Counter() unmapped = Counter() for record in human_records: name, method = ictv_species(record) if not name: unmapped[ncbi_species.get(record["virus tax id"]) or record["virus name"]] += 1 continue match_method[method] += 1 human_virus_taxa[name] += 1 if record["DISEASE"] or record["KEGG DISEASE"]: level = "disease" elif record["pmid"]: level = "literature" else: level = "annotation" if name not in tier or TIERS.index(level) < TIERS.index(tier[name]): tier[name] = level other_vertebrate = defaultdict(set) for record in records: if record["host tax id"] in ("", "1", HUMAN) or "Vertebrata" not in record["host lineage"]: continue name, _ = ictv_species(record) if name in tier: other_vertebrate[name].add(record["host name"]) bacterial = {name for name in tier if not vmr_has_vertebrate_host(vmr[name]["host"])} counted = sorted(set(tier) - bacterial) rows = [] for name in counted: entry = vmr[name] genome = "; ".join(sorted(entry["genome"])) rows.append({ "ictv_species": name, "realm": entry["realm"], "family": entry["family"], "genome": genome, "genome_group": genome_group(genome), "evidence_tier": tier[name], "human_virus_taxa": human_virus_taxa[name], "other_vertebrate_hosts": len(other_vertebrate[name]), "vmr_host_source": "; ".join(sorted(entry["host"])), }) args.output_dir.mkdir(parents=True, exist_ok=True) with open(args.output_dir / "human-host-species.csv", "w", encoding="utf-8", newline="") as stream: writer = csv.DictWriter(stream, fieldnames=list(rows[0]), lineterminator="\n") writer.writeheader() writer.writerows(rows) with open(args.output_dir / "unmapped-ncbi-names.csv", "w", encoding="utf-8", newline="") as stream: writer = csv.writer(stream, lineterminator="\n") writer.writerow(["ncbi_species_name", "human_host_records"]) writer.writerows(sorted(unmapped.items())) summary = [] def add(group, value, subset): subset = list(subset) summary.append({ "group": group, "value": value, "species": len(subset), **{tier_name: sum(row["evidence_tier"] == tier_name for row in subset) for tier_name in TIERS}, "also_other_vertebrate": sum(row["other_vertebrate_hosts"] > 0 for row in subset), }) add("all", "ICTV species with a recorded human host", rows) for group in ("genome_group", "realm"): for value, _ in Counter(row[group] for row in rows).most_common(): add(group, value, (row for row in rows if row[group] == value)) with open(args.output_dir / "summary.csv", "w", encoding="utf-8", newline="") as stream: writer = csv.DictWriter(stream, fieldnames=list(summary[0]), lineterminator="\n") writer.writeheader() writer.writerows(summary) report = { "inputs": {name: sha256(args.input_dir / name) for name in ("virushostdb.tsv", "virus_genome_type.tsv", VMR_FILE)}, "human_host_records": len(human_records), "human_host_ncbi_species": len({ncbi_species.get(r["virus tax id"]) for r in human_records}), "records_mapped": sum(match_method.values()), "ictv_species_with_human_host": len(tier), "excluded_no_vertebrate_host_species": sorted(bacterial), "counted_species": len(counted), "unmapped_ncbi_names": len(unmapped), "unmapped_human_host_records": sum(unmapped.values()), "vmr_species_total": len(vmr), } print(json.dumps(report, indent=2)) for line in summary: print(line) if __name__ == "__main__": main()