compare_v2.py

import sys
from Bio import SeqIO


def mismatch(seq1, seq2):
    index = []

    for i, (s1, s2) in enumerate(zip(seq1, seq2)):
        if s1 != s2:
            index.append(i)

    return index


def get_effect(seq, pos, ref, mismatches):
    effect = "."

    if len(mismatches) > 1:
        effect = "*"
    elif pos in mismatches:
        if seq.translate() == ref.translate():
            effect = "synonymous"
        else:
            effect = "non-synonymous"

    return effect


sequences = SeqIO.index(sys.argv[1], "fasta")
result = []

for i in range(0, len(sequences['species1'].seq), 3):
    codon_s1 = sequences['species1'].seq[i:i+3]
    codon_s2 = sequences['species2'].seq[i:i+3]
    codon_s3 = sequences['species3'].seq[i:i+3]

    if codon_s1 != codon_s2:
        mismatches_s1 = mismatch(codon_s1, codon_s3)
        mismatches_s2 = mismatch(codon_s2, codon_s3)

        pos = set(mismatches_s1 + mismatches_s2)

        for p in pos:
            effect_s1 = get_effect(codon_s1, p, codon_s3, mismatches_s1)
            effect_s2 = get_effect(codon_s2, p, codon_s3, mismatches_s2)

            result.append((
                i+p+1,
                codon_s3[p],
                codon_s1[p],
                effect_s1,
                codon_s2[p],
                effect_s2
            ))

print("#POS", "species3", "species1", "effect_s1", "species2", "effect_s2", sep="\t")

for row in result:
    print("\t".join(map(str, row)))

Second answer on biostars question: https://www.biostars.org/p/320742