Answer for SO (https://stackoverflow.com/q/69836764/246801)

chrM_final.csv

chrM_genes.csv

chrM_location.csv

main.py

import csv

from collections import defaultdict

# This will create a dictionary, where subject will be the key
# and the value will be a list (of rows)
chrM_rows_by_subject = defaultdict(list)


with open('chrM_genes.csv', newline='') as f:
    reader = csv.reader(f)
    next(reader)  # read (skip) header

    subject_col = 2
    for row in reader:
        # you mentioned empty rows, that divide subjects, so skip empty rows
        if row == []:
            continue

        subject = row[subject_col]
        chrM_rows_by_subject[subject].append(row)


# Return a row that matches `subject` with `pos` between start and end; or None.
def match_gene_row(subject, pos):
    rows = chrM_rows_by_subject[subject]

    for row in rows:
        start = row[3]
        end   = row[4]

        if pos >= start and pos <= end:
            # return just the data we want...
            return row

    # or return nothing at all
    return None


final_rows = []  # accumulate all rows here, for writing later

with open('chrM_location.csv', newline='') as f:
    reader = csv.reader(f)

    # Modify header
    header = next(reader)
    header.extend(['CDS','Start','End','sequence_cc'])
    final_rows.append(header)

    # Read rows and match to genes
    pos_column = 1
    for row in reader:
        pos = row[pos_column]
        matched_row = match_gene_row('CDS', pos)

        if matched_row is not None:
            subj, start, end = matched_row[2:5]
            computed = str(int(pos)-int(start)+1) # this is coming out negative??
            row.extend([subj, start, end, computed])
            final_rows.append(row)


# Finally, write
with open('final.csv', 'w', newline='') as f:
    writer = csv.writer(f)
    writer.writerows(final_rows)