A small module to join gene data tables from GenBank

join_table.py

# Version: 2022/12/07 - 10:51

import pandas as pd
import os
import sys

# Return all files in the given directory as a list
dir = "./tables"
files: list[str] = []
for file in os.listdir(dir):
    # List files that end with specific extension (.tab)
    if file.endswith(".tab"):
        files.append(os.path.join(dir, file))

# Exit if files list is empty
if not files:
    sys.exit(f"No table files in {dir}.")

# Iterate throw files and concatenate tables with Taxa and Lineages only
# no Accession Numbers for now.
table_list: list[pd.DataFrame] = []
for file in files:
    table = pd.read_table(file, header=None)[[0, 1]]
    table.columns = ["Taxa", "Lineage"]  # type: ignore
    table_list.append(table)

Taxa = pd.concat(table_list)

# Create a dictionary of all Lineages and Taxa attributed to them
names = Taxa.groupby('Lineage').Taxa.apply(  # type: ignore
    list).to_dict()

# Identify Taxa with more than one name and mark it with an *
for k, v in names.items():
    if len(set(v)) == 1:
        names[k] = v[0]
    else:
        names[k] = f"*{v[0]}"

# Turn names dict into a pandas dataframe
names = pd.DataFrame(list(names.items()), columns=["Lineage", "Taxa"])

# Merge dataframes and recover the Accession Numbers and other variables
# that may exist
for file in files:
    table = pd.read_table(file, header=None)[[1, 2]]
    table.columns = ["Lineage",
                     f"{os.path.basename(file.split('final', 1)[0])}"]  # type: ignore
    names = names.merge(table, how="outer", on="Lineage")

# Save final table with merged results as tab separated csv
names.to_csv('./JointTable.tab', sep="\t",  index=False)