What is the format for input using a csv file?

[DadongZ]

I have a large list of peptides and wondering what is the format if using a csv file as input. I have tried few ways but doesn't work. Is there a template?

[ikizhvatov]

Mentioning csv in the help is confusing, I see. Input csv should contain a peptide per line. I have added example_input.csv and the corresponding output csv to the repo.

[DadongZ]

Thanks! That's helpful.

[DadongZ]

Thanks, I got the below error with csv as input (~150K peptides):

Traceback (most recent call last):
  File "/home/dz33/.conda/envs/hmmhc/bin/hmmhc-predict", line 8, in <module>
    sys.exit(main())
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/hmmhc/cmdline.py", line 53, in main
    predictions = predictor.predict(peptides)
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/hmmhc/hmmhc.py", line 70, in predict
    normalizedLogOdds = self.computeLogOdds(peptidesList)
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/hmmhc/hmmhc.py", line 103, in computeLogOdds
    sequenceSetBlocks = self.toSequenceSetBlocks(peptideList)
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/hmmhc/hmmhc.py", line 156, in toSequenceSetBlocks
    [list(p) for p in peptideList[rangeStart:rangeEnd]]
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/ghmm.py", line 948, in __init__
    internalInput = [self.emissionDomain.internalSequence(seq) for seq in sequenceSetInput]
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/ghmm.py", line 393, in internalSequence
    result = map(lambda i: self.index[i], result)
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/ghmm.py", line 393, in <lambda>
    result = map(lambda i: self.index[i], result)
KeyError: 'X'

Some suggestions?

[ikizhvatov]

At least one of your peptide-encoding strings contains character 'X' which denotes an unknown amino acid and is not supported by the predictor. 'U' is also not supported.

As a quick solution, for now please input only the peptide strings containing the 20 amino acids ACDEFGHIKLMNPQRSTVWY.

We will consider filtering the peptides with unsupported amino acids in the tool itself.

[DadongZ]

I removed all peptides contains X and U but still got error

Traceback (most recent call last):
  File "/home/dz33/.conda/envs/hmmhc/bin/hmmhc-predict", line 8, in <module>
    sys.exit(main())
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/hmmhc/cmdline.py", line 53, in main
    predictions = predictor.predict(peptides)
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/hmmhc/hmmhc.py", line 70, in predict
    normalizedLogOdds = self.computeLogOdds(peptidesList)
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/hmmhc/hmmhc.py", line 103, in computeLogOdds
    sequenceSetBlocks = self.toSequenceSetBlocks(peptideList)
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/hmmhc/hmmhc.py", line 156, in toSequenceSetBlocks
    [list(p) for p in peptideList[rangeStart:rangeEnd]]
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/ghmm.py", line 948, in __init__
    internalInput = [self.emissionDomain.internalSequence(seq) for seq in sequenceSetInput]
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/ghmm.py", line 393, in internalSequence
    result = map(lambda i: self.index[i], result)
  File "/home/dz33/.conda/envs/hmmhc/lib/python2.7/site-packages/ghmm.py", line 393, in <lambda>
    result = map(lambda i: self.index[i], result)
KeyError: '*'

[ikizhvatov]

As said, peptide strings shall only contain the 20 canonical amino acids. You have '*' (it usually denotes a stop codon) in at least one of the strings.

[DadongZ]

A stop codon in peptide? These are peptides from genes CDS though and I have removed all X/U letters.