Yeast gene symbols to human homologs

yeast2human.py

import numpy as np
import pandas as pd
import requests

# Interested genes
df = pd.read_excel("yeast_genes.xlsx")
pom_genes = df["Schizosaccharomyces pombe"][
    df["Schizosaccharomyces pombe"].notnull()
].tolist()
cer_genes = df["Saccharomyces cerevisiae"][
    df["Saccharomyces cerevisiae"].notnull()
].tolist()

# SGD data
cer_map = pd.read_table(
    "http://sgd-archive.yeastgenome.org/curation/chromosomal_feature/SGD_features.tab",
    header=None,
)
cer_map = cer_map.iloc[:, [0, 4]]
cer_map.columns = ["SGDID", "Symbol"]
cer_map = cer_map[cer_map["Symbol"].isin(cer_genes)].reset_index(drop=True)

# Get human homologs from Alliance of Genome Resources
ss = requests.session()
payload = {
    "filter.stringency": "stringent",
    "taxonID": "NCBITaxon:9606",
}
res = []
for i, row in cer_map.iterrows():
    r = ss.get(
        f"https://www.alliancegenome.org/api/gene/SGD:{row['SGDID']}/homologs",
        params=payload,
    )
    homologs = [
        x["homologGene"]["symbol"]
        for x in r.json()["results"]
        if x["homologGene"]["species"]["taxonId"] == "NCBITaxon:9606"
    ]
    res.append(homologs)

cer_map["HumanSymbol"] = res
cer_map = cer_map[["Symbol", "HumanSymbol"]]

# Fission yeast (Schizosaccharomyces pombe) data from PomBase
pom_ortho = pd.read_table(
    "ftp://ftp.pombase.org/pombe/orthologs/human-orthologs.txt.gz", header=None,
)
pom_ortho.columns = ["Systematic", "HumanSymbol"]
pom_ortho = (
    pom_ortho.assign(HumanSymbol=pom_ortho["HumanSymbol"].str.split("|"))
    .explode("HumanSymbol")
    .query("(HumanSymbol.notnull()) & (HumanSymbol != 'NONE')")
)

pom_map = pd.read_table(
    "ftp://ftp.pombase.org/pombe/names_and_identifiers/gene_IDs_names.tsv",
    skiprows=1,
    header=None,
)
pom_map.columns = ["Systematic", "Symbol", "Synonyms"]
pom_map = (
    pom_map.drop(["Synonyms"], axis=1)
    .assign(Symbol=pom_map["Symbol"].str.upper())
    .query("Symbol.isin(@pom_genes)")
    .merge(pom_ortho, on="Systematic")
    .groupby(["Systematic", "Symbol"])["HumanSymbol"]
    .apply(list)
    .reset_index(name="HumanSymbol")[["Symbol", "HumanSymbol"]]
)

# Merge results
res = (
    df.merge(cer_map, how="left", left_on="Saccharomyces cerevisiae", right_on="Symbol")
    .drop("Symbol", axis=1)
    .merge(pom_map, how="left", left_on="Schizosaccharomyces pombe", right_on="Symbol")
    .drop("Symbol", axis=1)
)

res = res.assign(
    HumanSymbol_x=res["HumanSymbol_x"].apply(lambda x: [] if x is np.NaN else x)
).assign(HumanSymbol_y=res["HumanSymbol_y"].apply(lambda x: [] if x is np.NaN else x))

res = res.assign(HumanSymbol=res["HumanSymbol_x"] + res["HumanSymbol_y"]).drop(
    ["HumanSymbol_x", "HumanSymbol_y"], axis=1
)

res["HumanSymbol"] = res["HumanSymbol"].apply(lambda x: sorted(list(set(x))))
res.rename(columns={"HumanSymbol": "Homo sapiens"}).to_excel(
    "yeast_table_1.xlsx", index=False
)