This is a test version of Biostars. For the public version, visit https://www.biostars.org.
pairwise2 extended_protein alphabet

I am getting a Key Error when I try to use pairwise2 alignment. I have a translated generic_protein sequence and an extended_protein sequence from an indexed fasta file.

File "/biopython-1.70/Bio/pairwise2.py", line 407, in __call__
    return _align(**keywds)
  File "/biopython-1.70/Bio/pairwise2.py", line 457, in _align
    align_globally, score_only)
  File "/biopython-1.70/Bio/pairwise2.py", line 649, in _make_score_matrix_fast
    match_fn(sequenceA[row - 1], sequenceB[col - 1])
  File "/biopython-1.70/Bio/pairwise2.py", line 982, in __call__
    return self.score_dict[(charA, charB)]
KeyError: ('U', 'M')

Can pairwise2 do an alignment with the IUPAC Extended Protein Alphabet?

Example sequences that return the error above:

Translated generic_protein: 
MYISPEEFKPIAEKLTGSVPVANYEEEELSHDPSEETVTIESRFQPLLTETMTKSKDGFLGVSRLALSGLRNWTTAASPSAVFAARHFRPFLPPPGQELGQPWWIIPGELSVFTGYLSNNRFYPPPPKGKEVRTALGDRGCTWPQLESGPSCVNLDPST

Indexed extended_protein sequence:
MPSVFQKNIISRVSQIQLVPTGTTPPAKMPEDTAENWPGLVIMATTEHKPRFNYGGGGGGGGGTPLLLSGSSCPLVRTFGTGIGRLGPDSCLLAQPLPPKQESALKVLGTDGLFLFSSLDTDQDMYISPEEFKPIAEKLTGSVPEASYEEEELHHDPSEETLTIEARFQPLLTETMTKSKDGFLGVSRLALSGLRNWTTAASPTAEFAARHFRAFLPPPGQELGQPWWIIPGELSVFTGYLSNNRFYPPPPKDKEVIIHRLLSMFHPRPFVKTRFAPQGTVACLTAISDSYYTVMFRIHAEFQLSEPPDFPFWFSPGQFTGHIILSKDATHIRDFRLFVPNHRSLNVDMEWLYGASETSNMEVDIGYIPQMELEAMGPSVPSVILDEDGNMIDSRLPSGEPLQFVFEEIKWHQELSWEEAARRLEVAMYPFKKVSYLPFTEAFDRAKAENKLVHSILLWGALDDQSCUGSGRTLRETVLESPPILTLLNESFISTWSLVKELEDLQSQQENALHRQLAGLHLEKYSFPVEMMVCLPNGTVVHHINANYFLDITSMKPEDLDNKVFSFSSTFEDPSTATYMQFLREGLRRGLPLLEP
alignment sequence biopython python pairwise2
• 2,480 views • link
updated by Andrzej Zielezinski Andrzej Zielezinski • written by aocken • aocken

1 answer

The problem is that matrix used for amino acid substitutions does not include extra amino acids (e.g. U). And this causes further errors. However, you can easily overcome this issue by creating a callback function that will handle such cases.

Python code:

from Bio import pairwise2
from Bio.pairwise2 import format_alignment
from Bio.SubsMat import MatrixInfo as matlist

seq1 = 'KAENKLVHSIILWSALDQSCUGSGRTLRETVLESP'        # There is U in it.
seq2 = 'KAENKLVHSILLWGALDDQSCGSGRTLRETVLESP'

matrix = matlist.blosum62

# Callback function
def matrix_get(aa1, aa2):
    if (aa1, aa2) in matrix:
        val = matrix[(aa1, aa2)]
    elif (aa2, aa1) in matrix:
        val = matrix[(aa2, aa1)]
    else:
        val = 0
    return val

for a in pairwise2.align.globalcs(seq1, seq2, matrix_get, -10, -0.5):
    print(format_alignment(*a))

Output:

KAENKLVHSIILWSALD-QSCUGSGRTLRETVLESP
||||||||||||||||||||||||||||||||||||
KAENKLVHSILLWGALDDQSC-GSGRTLRETVLESP
  Score=157

KAENKLVHSIILWSAL-DQSCUGSGRTLRETVLESP
||||||||||||||||||||||||||||||||||||
KAENKLVHSILLWGALDDQSC-GSGRTLRETVLESP
  Score=157
• 0 views • link
written by Andrzej Zielezinski Andrzej Zielezinski

Log in to answer this question.

More posts like this