Mercurial > repos > yufei-luo > s_mart
view commons/core/coord/Match.py @ 47:b6481845eb0d
Uploaded
author | m-zytnicki |
---|---|
date | Mon, 30 Sep 2013 05:51:28 -0400 |
parents | 769e306b7933 |
children |
line wrap: on
line source
# Copyright INRA (Institut National de la Recherche Agronomique) # http://www.inra.fr # http://urgi.versailles.inra.fr # # This software is governed by the CeCILL license under French law and # abiding by the rules of distribution of free software. You can use, # modify and/ or redistribute the software under the terms of the CeCILL # license as circulated by CEA, CNRS and INRIA at the following URL # "http://www.cecill.info". # # As a counterpart to the access to the source code and rights to copy, # modify and redistribute granted by the license, users are provided only # with a limited warranty and the software's author, the holder of the # economic rights, and the successive licensors have only limited # liability. # # In this respect, the user's attention is drawn to the risks associated # with loading, using, modifying and/or developing or reproducing the # software by the user in light of its specific status of free software, # that may mean that it is complicated to manipulate, and that also # therefore means that it is reserved for developers and experienced # professionals having in-depth computer knowledge. Users are therefore # encouraged to load and test the software's suitability as regards their # requirements in conditions enabling the security of their systems and/or # data to be ensured and, more generally, to use and operate it in the # same conditions as regards security. # # The fact that you are presently reading this means that you have had # knowledge of the CeCILL license and that you accept its terms. import sys from commons.core.coord.Range import Range from commons.core.coord.Path import Path ## Handle a chain of match(es) between two sequences, query and subject, with an identifier and the length of the input sequences # class Match( Path ): ## Constructor # def __init__(self): Path.__init__(self) self.query_length = -1 self.query_length_perc = -1 # length of the match on the query / length of the query self.query_seqlength = -1 self.match_length_perc = -1 # length of the match on the query / total length of the subject self.subject_length = -1 self.subject_length_perc = -1 # length of the match on the subject / length of the subject self.subject_seqlength = -1 ## Equal operator # def __eq__(self, o): if o == None \ or self.query_length != o.query_length or self.query_length_perc != o.query_length_perc\ or self.query_seqlength != o.query_seqlength or self.subject_length != o.subject_length\ or self.subject_length_perc != o.subject_length_perc or self.subject_seqlength != o.subject_seqlength\ or self.match_length_perc != o.match_length_perc: return False return Path.__eq__(self, o) ## Return the length of the match on the query divided by the total length of the query # def getLengthPercOnQuery(self): return self.query_length_perc ## Return the length of the match on the subject divided by the total length of the subject # def getLengthPercOnSubject(self): return self.subject_length_perc ## Return the length of the match on the subject # def getLengthMatchOnSubject(self): return self.subject_length ## Set attributes from a tuple # # @param tuple: a tuple with (query name,query start,query end, # query length, query length perc (between 0-1), match length perc (between 0-1), subject name, # subject start,subject end,subject length, subject length percentage (between 0-1), e_value,score,identity,id) # def setFromTuple( self, tuple ): queryStart = int(tuple[1]) queryEnd = int(tuple[2]) subjectStart = int(tuple[7]) subjectEnd = int(tuple[8]) if queryStart < queryEnd: self.range_query = Range(tuple[0],queryStart,queryEnd) self.range_subject = Range(tuple[6],subjectStart,subjectEnd) else: self.range_query = Range(tuple[0],queryEnd,queryStart) self.range_subject = Range(tuple[6],subjectEnd,subjectStart) self.query_length = int(tuple[3]) self.query_length_perc = float(tuple[4]) self.query_seqlength = int( self.query_length / self.query_length_perc ) self.match_length_perc = float(tuple[5]) self.subject_length = int(tuple[9]) self.subject_length_perc = float(tuple[10]) self.subject_seqlength = int( self.subject_length / self.subject_length_perc ) self.e_value = float(tuple[11]) self.score = float(tuple[12]) self.identity = float(tuple[13]) self.id = int(tuple[14]) ## Reset # def reset( self ): Path.reset( self ) self.query_length = -1 self.query_length_perc = -1 self.query_seqlength = -1 self.match_length_perc = -1 self.subject_length = -1 self.subject_length_perc = -1 self.subject_seqlength = -1 ## Return a formated string of the attribute data # def toString( self ): string = "%s" % ( self.range_query.toString() ) string += "\t%i\t%f" % ( self.query_length, self.query_length_perc ) string += "\t%f" % ( self.match_length_perc ) string += "\t%s" % ( self.range_subject.toString() ) string += "\t%i\t%f" % ( self.subject_length, self.subject_length_perc ) string += "\t%g\t%i\t%f" % ( self.e_value, self.score, self.identity ) string += "\t%i" % ( self.id ) return string ## Return a Path instance # def getPathInstance( self ): p = Path() tuple = ( self.id, self.range_query.seqname, self.range_query.start, self.range_query.end, self.range_subject.seqname, self.range_subject.start, self.range_subject.end, self.e_value, self.score, self.identity ) p.setFromTuple( tuple ) return p ## Give information about a match whose query is included in the subject # # @return string # def getQryIsIncluded( self ): string = "query %s (%d bp: %d-%d) is contained in subject %s (%d bp: %d-%d): id=%.2f - %.3f - %.3f - %.3f" %\ ( self.range_query.seqname, self.query_seqlength, self.range_query.start, self.range_query.end, self.range_subject.seqname, self.subject_seqlength, self.range_subject.start, self.range_subject.end, self.identity, self.query_length_perc, self.match_length_perc, self.subject_length_perc ) return string def increaseLengthPercOnQuery(self, coverage): self.query_length_perc += coverage ## Compare the object with another match and see if they are equal # (same identity, E-value and score + same subsequences whether in query or subject) # # @return True if objects are equals False otherwise # def isDoublonWith( self, match, verbose=0 ): # if both matches have same identity, score and E-value if self.identity == match.identity and self.score == match.score and self.e_value == match.e_value: # if query and subject are identical if ( self.range_query.seqname == match.range_query.seqname \ and self.range_subject.seqname == match.range_subject.seqname ): # if the coordinates are equal if self.range_query.__eq__( match.range_query ) and self.range_subject.__eq__( match.range_subject ): return True else: if verbose > 0: print "different coordinates"; sys.stdout.flush() return False # if query and subject are reversed but identical elif self.range_query.seqname == match.range_subject.seqname and self.range_subject.seqname == match.range_query.seqname: # if the coordinates are equal if self.range_query.__eq__( match.range_subject ) and self.range_subject.__eq__( match.range_query ): return True else: if verbose > 0: print "different coordinates"; sys.stdout.flush() return False else: if verbose > 0: print "different sequence names"; sys.stdout.flush() return False else: if verbose > 0: print "different match numbers"; sys.stdout.flush() return False