Added a clustal test file, an example notebook, and alignment parsing… #1
Conversation
… in the _read function. Also made two hidden funcitons to iterate sequence and alignment biopython objects.
|
Awesome! This looks good. I'm testing it now and will leave comments if I see things that need fixing. |
|
One general comment about coding style: I typically use a "lowercase and underscore" convention when naming instance variables, functions, This convention was set by PEP 8. I'd like to follow PEP 8 as best as we can. I'll leave comments in the PR where this convention is broken. |
|
|
||
| # Distinguish sequence vs. alignment, pass SeqRecords | ||
| if schema == 'clustal': | ||
| alignIter = AlignIO.parse(filename, format=schema, **kwargs) |
There was a problem hiding this comment.
follow PEP 8: alignIter -> align_iter
| if schema == 'clustal': | ||
| alignIter = AlignIO.parse(filename, format=schema, **kwargs) | ||
| # Parse the multiple sequence alignments | ||
| alignDict = _parseAlignRec(alignIter, seq_label) |
There was a problem hiding this comment.
follow PEP 8: alignDict -> align_dict
| # Return DataFrame. | ||
| return data | ||
|
|
||
| def _parseAlignRec(alignIter, seq_label): |
There was a problem hiding this comment.
follow PEP 8:
_parseAlignRec->_parse_align_recordalignIter->align_iter
|
|
||
| return alignData | ||
|
|
||
| def _parseSeqRec(seqIter, seq_label): |
There was a problem hiding this comment.
follow PEP 8:
_parseSeqRec->_parse_seq_recseqIter->seq_iter
| def _parseSeqRec(seqIter, seq_label): | ||
| """A Bio.SequenceRecord parser. | ||
| """ | ||
| seqDict = dict() |
There was a problem hiding this comment.
follow PEP 8: seqDict -> seq_dict
Continue below.
I think I agree with this statement... give each item a value in the Could you explain the output in the example clustal? Are each mouse/opossum line-pairs just segments of one longer sequence? Or are each pair a different sequence? |
… in the _read function. Also made two hidden funcitons to iterate sequence and alignment biopython objects.
Did a bit of hacking to get the dataframes to have the same form as previously. Multiple sequence alignments are spit out as a multiindexed dataframe.