commit: 74035c984818c6829c1f82cbcc1f419ff45f93d0 Author: Martin Mokrejš <mmokrejs <AT> fold <DOT> natur <DOT> cuni <DOT> cz> AuthorDate: Fri Apr 18 18:21:20 2014 +0000 Commit: Martin Mokrejs <mmokrejs <AT> fold <DOT> natur <DOT> cuni <DOT> cz> CommitDate: Fri Apr 18 18:21:20 2014 +0000 URL: http://git.overlays.gentoo.org/gitweb/?p=proj/sci.git;a=commit;h=74035c98
sci-biology/biopython: ops, taking back my own testing patches Package-Manager: portage-2.2.7 --- sci-biology/biopython/ChangeLog | 4 + sci-biology/biopython/files/SeqRecord.py.patch | 148 --------------------- .../biopython/files/adjust-trimpoints.patch | 76 ----------- 3 files changed, 4 insertions(+), 224 deletions(-) diff --git a/sci-biology/biopython/ChangeLog b/sci-biology/biopython/ChangeLog index c326c4a..796f6a9 100644 --- a/sci-biology/biopython/ChangeLog +++ b/sci-biology/biopython/ChangeLog @@ -2,6 +2,10 @@ # Copyright 1999-2014 Gentoo Foundation; Distributed under the GPL v2 # $Header: $ + 18 Apr 2014; Martin Mokrejs <[email protected]> + -files/SeqRecord.py.patch, -files/adjust-trimpoints.patch: + sci-biology/biopython: ops, taking back my own testing patches + *biopython-1.62-r3 (18 Apr 2014) *biopython-1.62-r4 (18 Apr 2014) diff --git a/sci-biology/biopython/files/SeqRecord.py.patch b/sci-biology/biopython/files/SeqRecord.py.patch deleted file mode 100644 index ac3785f..0000000 --- a/sci-biology/biopython/files/SeqRecord.py.patch +++ /dev/null @@ -1,148 +0,0 @@ -diff --git a/Bio/SeqIO/SffIO.py b/Bio/SeqIO/SffIO.py -index 1971dba..43b38fd 100644 ---- a/Bio/SeqIO/SffIO.py -+++ b/Bio/SeqIO/SffIO.py -@@ -539,8 +539,15 @@ _valid_UAN_read_name = re.compile(r'^[a-zA-Z0-9]{14}$') - - - def _sff_read_seq_record(handle, number_of_flows_per_read, flow_chars, -- key_sequence, alphabet, trim=False): -- """Parse the next read in the file, return data as a SeqRecord (PRIVATE).""" -+ key_sequence, alphabet, trim=False, interpret_qual_trims=True, interpret_adapter_trims=False): -+ """Parse the next read in the file, return data as a SeqRecord (PRIVATE). -+ Allow user to specify which type of clipping values should be applied -+ while reading the SFF stream. To be backwards compatible, we interpret -+ only the quality-based trim points by default. That results in lower-cased -+ sequences in the low-qual region, regardless what adapter-based clip points -+ say. This should be the desired behavior. More discussion at -+ https://redmine.open-bio.org/issues/3437 -+ """ - #Now on to the reads... - #the read header format (fixed part): - #read_header_length H -@@ -589,20 +596,41 @@ def _sff_read_seq_record(handle, number_of_flows_per_read, flow_chars, - warnings.warn("Post quality %i byte padding region contained data, SFF data is not broken" - % padding) - #Follow Roche and apply most aggressive of qual and adapter clipping. -- #Note Roche seems to ignore adapter clip fields when writing SFF, -- #and uses just the quality clipping values for any clipping. -- clip_left = max(clip_qual_left, clip_adapter_left) -- #Right clipping of zero means no clipping -- if clip_qual_right: -- if clip_adapter_right: -- clip_right = min(clip_qual_right, clip_adapter_right) -+ #Note Roche does not use adapter clip fields when writing SFF files -+ #but instead combines the adapter clipping information with quality-based -+ #values and writes the most aggressive combination into clip fields (as -+ #allowed by SFF specs). -+ -+ if interpret_qual_trims: -+ if interpret_adapter_trims: -+ clip_left = max(clip_qual_left, clip_adapter_left) -+ #Right clipping of zero means no clipping -+ if clip_qual_right: -+ if clip_adapter_right: -+ clip_right = min(clip_qual_right, clip_adapter_right) -+ else: -+ #Typical case with Roche SFF files -+ clip_right = clip_qual_right -+ elif clip_adapter_right: -+ clip_right = clip_adapter_right -+ else: -+ clip_right = seq_len - else: -- #Typical case with Roche SFF files -- clip_right = clip_qual_right -- elif clip_adapter_right: -- clip_right = clip_adapter_right -+ clip_left = clip_qual_left -+ if clip_qual_right: -+ clip_right = clip_qual_right -+ else: -+ clip_right = seq_len -+ elif interpret_adapter_trims: -+ clip_left = clip_adapter_left -+ if clip_adapter_right: -+ clip_right = clip_adapter_right -+ else: -+ clip_right = seq_len - else: -- clip_right = seq_len -+ clip_left = 0 -+ clip_right = seq_len -+ - #Now build a SeqRecord - if trim: - seq = seq[clip_left:clip_right].upper() -diff --git a/Bio/SeqRecord.py b/Bio/SeqRecord.py -index c90e13b..66bdea0 100644 ---- a/Bio/SeqRecord.py -+++ b/Bio/SeqRecord.py -@@ -14,6 +14,8 @@ __docformat__ = "epytext en" # Simple markup to show doctests nicely - # also BioSQL.BioSeq.DBSeq which is the "Database Seq" class) - - -+from Bio.Seq import Seq -+ - class _RestrictedDict(dict): - """Dict which only allows sequences of given length as values (PRIVATE). - -@@ -76,7 +78,7 @@ class _RestrictedDict(dict): - if not hasattr(value, "__len__") or not hasattr(value, "__getitem__") \ - or (hasattr(self, "_length") and len(value) != self._length): - raise TypeError("We only allow python sequences (lists, tuples or " -- "strings) of length %i." % self._length) -+ "strings) of length %i whereas you passed an object of length %s." % (self._length, str(len(value)))) - dict.__setitem__(self, key, value) - - def update(self, new_dict): -@@ -290,10 +292,11 @@ class SeqRecord(object): - """) - - def _set_seq(self, value): -- #TODO - Add a deprecation warning that the seq should be write only? -- if self._per_letter_annotations: -- #TODO - Make this a warning? Silently empty the dictionary? -- raise ValueError("You must empty the letter annotations first!") -+ # we should be much more user friendly and accept even a plain sequence string -+ # and make the Seq or MutableSeq object ourselves -+ if not isinstance(value, Seq): -+ raise ValueError("You must pass a Seq object containing the new sequence instead of just plain string.") -+ else: - self._seq = value - try: - self._per_letter_annotations = _RestrictedDict(length=len(self.seq)) -@@ -696,7 +699,7 @@ class SeqRecord(object): - SeqIO.write(self, handle, format_spec) - return handle.getvalue() - -- def __len__(self): -+ def __len__(self, trim=False, interpret_qual_trims=True, interpret_adapter_trims=False): - """Returns the length of the sequence. - - For example, using Bio.SeqIO to read in a FASTA nucleotide file: -@@ -707,6 +710,10 @@ class SeqRecord(object): - 309 - >>> len(record.seq) - 309 -+ -+ It should be possible to get length of a raw object, of trimmed -+ object by quality or adapter criteria or both, whenever user wants -+ to, not only when data is parsed from input. - """ - return len(self.seq) - -@@ -725,6 +732,13 @@ class SeqRecord(object): - """ - return True - -+ def apply_trimpoints(self, trim=False, interpret_qual_trims=False, interpret_adapter_trims=False): -+ """We should apply either of the quality-based or adapter-based annotated -+ trim points and return a new, sliced object. -+ """ -+ pass -+ -+ - def __add__(self, other): - """Add another sequence or string to this sequence. - diff --git a/sci-biology/biopython/files/adjust-trimpoints.patch b/sci-biology/biopython/files/adjust-trimpoints.patch deleted file mode 100644 index dd6d548..0000000 --- a/sci-biology/biopython/files/adjust-trimpoints.patch +++ /dev/null @@ -1,76 +0,0 @@ -diff --git a/Bio/SeqIO/SffIO.py b/Bio/SeqIO/SffIO.py -index 1971dba..43b38fd 100644 ---- a/Bio/SeqIO/SffIO.py -+++ b/Bio/SeqIO/SffIO.py -@@ -539,8 +539,15 @@ _valid_UAN_read_name = re.compile(r'^[a-zA-Z0-9]{14}$') - - - def _sff_read_seq_record(handle, number_of_flows_per_read, flow_chars, -- key_sequence, alphabet, trim=False): -- """Parse the next read in the file, return data as a SeqRecord (PRIVATE).""" -+ key_sequence, alphabet, trim=False, interpret_qual_trims=True, interpret_adapter_trims=False): -+ """Parse the next read in the file, return data as a SeqRecord (PRIVATE). -+ Allow user to specify which type of clipping values should be applied -+ while reading the SFF stream. To be backwards compatible, we interpret -+ only the quality-based trim points by default. That results in lower-cased -+ sequences in the low-qual region, regardless what adapter-based clip points -+ say. This should be the desired behavior. More discussion at -+ https://redmine.open-bio.org/issues/3437 -+ """ - #Now on to the reads... - #the read header format (fixed part): - #read_header_length H -@@ -589,20 +596,41 @@ def _sff_read_seq_record(handle, number_of_flows_per_read, flow_chars, - warnings.warn("Post quality %i byte padding region contained data, SFF data is not broken" - % padding) - #Follow Roche and apply most aggressive of qual and adapter clipping. -- #Note Roche seems to ignore adapter clip fields when writing SFF, -- #and uses just the quality clipping values for any clipping. -- clip_left = max(clip_qual_left, clip_adapter_left) -- #Right clipping of zero means no clipping -- if clip_qual_right: -- if clip_adapter_right: -- clip_right = min(clip_qual_right, clip_adapter_right) -+ #Note Roche does not use adapter clip fields when writing SFF files -+ #but instead combines the adapter clipping information with quality-based -+ #values and writes the most aggressive combination into clip fields (as -+ #allowed by SFF specs). -+ -+ if interpret_qual_trims: -+ if interpret_adapter_trims: -+ clip_left = max(clip_qual_left, clip_adapter_left) -+ #Right clipping of zero means no clipping -+ if clip_qual_right: -+ if clip_adapter_right: -+ clip_right = min(clip_qual_right, clip_adapter_right) -+ else: -+ #Typical case with Roche SFF files -+ clip_right = clip_qual_right -+ elif clip_adapter_right: -+ clip_right = clip_adapter_right -+ else: -+ clip_right = seq_len - else: -- #Typical case with Roche SFF files -- clip_right = clip_qual_right -- elif clip_adapter_right: -- clip_right = clip_adapter_right -+ clip_left = clip_qual_left -+ if clip_qual_right: -+ clip_right = clip_qual_right -+ else: -+ clip_right = seq_len -+ elif interpret_adapter_trims: -+ clip_left = clip_adapter_left -+ if clip_adapter_right: -+ clip_right = clip_adapter_right -+ else: -+ clip_right = seq_len - else: -- clip_right = seq_len -+ clip_left = 0 -+ clip_right = seq_len -+ - #Now build a SeqRecord - if trim: - seq = seq[clip_left:clip_right].upper()
