comparison defuse_trinity_analysis.py @ 36:4353f776dfa3

Add defuse_trinity_analysis

36:4353f776dfa3

#!/usr/bin/env python
"""
#
#------------------------------------------------------------------------------
#                         University of Minnesota
#         Copyright 2014, Regents of the University of Minnesota
#------------------------------------------------------------------------------
# Author:
#
#  James E Johnson
#
#------------------------------------------------------------------------------
"""


"""
This tool takes the defuse results.tsv  tab-delimited file, trinity 
and creates a tabular report
"""

import sys,re,os.path
import optparse
from optparse import OptionParser

revcompl = lambda x: ''.join([{'A':'T','C':'G','G':'C','T':'A','a':'t','c':'g','g':'c','t':'a','N':'N','n':'n'}[B] for B in x][::-1])

def read_fasta(fp):
    name, seq = None, []
    for line in fp:
        line = line.rstrip()
        if line.startswith(">"):
            if name: yield (name, ''.join(seq))
            name, seq = line, []
        else:
            seq.append(line)
    if name: yield (name, ''.join(seq))


def test_rcomplement(seq, target):
  try:
    comp = revcompl(seq)
    return comp in target
  except:
    pass
  return False

def test_reverse(seq,target):
  return options.test_reverse and seq and seq[::-1] in target

def cmp_alphanumeric(s1,s2):
  if s1 == s2:
    return 0
  a1 = re.findall("\d+|[a-zA-Z]+",s1)
  a2 = re.findall("\d+|[a-zA-Z]+",s2)
  for i in range(min(len(a1),len(a2))):
    if a1[i] == a2[i]:
      continue
    if a1[i].isdigit() and a2[i].isdigit():
      return int(a1[i]) - int(a2[i])
    return 1 if a1[i] >  a2[i] else -1
  return len(a1) - len(a2)


def parse_defuse_results(inputFile): 
  columns = []
  defuse_results = []
  # {cluster_id : { field : value})
  try:
    for linenum,line in enumerate(inputFile):
      ## print >> sys.stderr, "%d: %s\n" % (linenum,line)
      fields = line.strip().split('\t')
      if line.startswith('cluster_id'):
        columns = fields
        ## print >> sys.stderr, "columns: %s\n" % columns
        continue
      cluster_dict = dict()
      cluster_id = fields[columns.index('cluster_id')]
      cluster_dict['cluster_id'] = fields[columns.index('cluster_id')]
      cluster_dict['gene_chromosome1'] = fields[columns.index('gene_chromosome1')]
      cluster_dict['gene_chromosome2'] = fields[columns.index('gene_chromosome2')]
      cluster_dict['genomic_strand1'] = fields[columns.index('genomic_strand1')]
      cluster_dict['genomic_strand2'] = fields[columns.index('genomic_strand2')]
      cluster_dict['gene1'] = fields[columns.index('gene1')]
      cluster_dict['gene2'] = fields[columns.index('gene2')]
      cluster_dict['gene_name1'] = fields[columns.index('gene_name1')]
      cluster_dict['gene_name2'] = fields[columns.index('gene_name2')]
      cluster_dict['gene_location1'] = fields[columns.index('gene_location1')]
      cluster_dict['gene_location2'] = fields[columns.index('gene_location2')]
      cluster_dict['expression1'] = int(fields[columns.index('expression1')])
      cluster_dict['expression2'] = int(fields[columns.index('expression2')])
      cluster_dict['genomic_break_pos1'] = int(fields[columns.index('genomic_break_pos1')])
      cluster_dict['genomic_break_pos2'] = int(fields[columns.index('genomic_break_pos2')])
      cluster_dict['breakpoint_homology'] = int(fields[columns.index('breakpoint_homology')])
      cluster_dict['orf'] = fields[columns.index('orf')] == 'Y'
      cluster_dict['exonboundaries'] = fields[columns.index('exonboundaries')] == 'Y'
      cluster_dict['read_through'] = fields[columns.index('read_through')] == 'Y'
      cluster_dict['interchromosomal'] = fields[columns.index('interchromosomal')] == 'Y'
      cluster_dict['adjacent'] = fields[columns.index('adjacent')] == 'Y'
      cluster_dict['altsplice'] = fields[columns.index('altsplice')] == 'Y'
      cluster_dict['deletion'] = fields[columns.index('deletion')] == 'Y'
      cluster_dict['eversion'] = fields[columns.index('eversion')] == 'Y'
      cluster_dict['inversion'] = fields[columns.index('inversion')] == 'Y'
      cluster_dict['span_count'] = int(fields[columns.index('span_count')])
      cluster_dict['splitr_count'] = int(fields[columns.index('splitr_count')])
      cluster_dict['splice_score'] = int(fields[columns.index('splice_score')])
      cluster_dict['probability'] = float(fields[columns.index('probability')] if columns.index('probability') else 'nan')
      cluster_dict['splitr_sequence'] = fields[columns.index('splitr_sequence')]
      defuse_results.append(cluster_dict) 
  except Exception, e:
    print >> sys.stderr, "failed: %s" % e
    sys.exit(1)
  return defuse_results

## deFuse params to the mapping application?

def __main__():
  #Parse Command Line
  parser = optparse.OptionParser()
  # files
  parser.add_option( '-i', '--input', dest='input', help='The input defuse results.tsv file (else read from stdin)' )
  parser.add_option( '-t', '--transcripts', dest='transcripts', default=None, help='Trinity transcripts' )
  parser.add_option( '-p', '--peptides', dest='peptides', default=None, help='Trinity ORFs' )
  parser.add_option( '-o', '--output', dest='output', help='The output report (else write to stdout)' )
  parser.add_option( '-a', '--transcript_alignment', dest='transcript_alignment', help='The output alignment file' )
  parser.add_option( '-A', '--orf_alignment', dest='orf_alignment', help='The output alignment file' )
  parser.add_option( '-N', '--nbases', dest='nbases', type='int', default=12, help='Number of bases on either side of the fusion to compare' )
  parser.add_option( '-L', '--min_pep_len', dest='min_pep_len', type='int', default=100, help='Minimum length of peptide to report' )
  parser.add_option( '-T', '--ticdist', dest='ticdist', type='int', default=1000000, help='Maximum intrachromosomal distance to be classified a Transcription-induced chimera (TIC)' )
  parser.add_option( '-P', '--prior_aa', dest='prior_aa', type='int', default=11, help='Number of protein AAs to show preceeding fusion point' )
  # min_orf_len
  # split_na_len
  # tic_len = 1000000
  # prior
  # deFuse direction reversed 
  # in frame ?
  # contain known protein elements
  # what protein change
  # trinity provides full transctipt, defuse doesn't show full
  #parser.add_option( '-r', '--reference', dest='reference', default=None, help='The genomic reference fasta' )
  #parser.add_option( '-g', '--gtf', dest='gtf', default=None, help='The genomic reference gtf feature file')
  (options, args) = parser.parse_args()

  # results.tsv input 
  if options.input != None:
    try:
      inputPath = os.path.abspath(options.input)
      inputFile = open(inputPath, 'r')
    except Exception, e:
      print >> sys.stderr, "failed: %s" % e
      exit(2)
  else:
    inputFile = sys.stdin
  # vcf output 
  if options.output != None:
    try:
      outputPath = os.path.abspath(options.output)
      outputFile = open(outputPath, 'w')
    except Exception, e:
      print >> sys.stderr, "failed: %s" % e
      exit(3)
  else:
    outputFile = sys.stdout

  ## Read defuse results
  fusions = parse_defuse_results(inputFile)
  ## Create a field with the 12 nt before and after the fusion point. 
  ## Create a field with the reverse complement of the 24 nt fusion point field.
  ## Add fusion type filed (INTER, INTRA, TIC)
  for i,fusion in enumerate(fusions):
      fusion['ordinal'] = i + 1
      split_seqs = fusion['splitr_sequence'].split('|')
      fusion['split_seqs'] = split_seqs
      fwd_seq = split_seqs[0][-(min(abs(options.nbases),len(split_seqs[0]))):] + split_seqs[1][:min(abs(options.nbases),len(split_seqs[1]))]
      rev_seq =  revcompl(fwd_seq)
      fusion['fwd_seq'] = fwd_seq
      fusion['rev_seq'] = rev_seq
      fusion_type = 'inter' if fusion['gene_chromosome1'] != fusion['gene_chromosome2'] else 'intra' if abs(fusion['genomic_break_pos1'] - fusion['genomic_break_pos2']) > options.ticdist else 'TIC'
      fusion['fusion_type'] = fusion_type
      fusion['transcripts'] = []
      fusion['Transcript'] = 'No'
      fusion['Protein'] = 'No'
      print >> sys.stdout, "%4d\t%6s\t%s\t%s\t%s\t%s\t%s" % (i,fusion['cluster_id'],fwd_seq,rev_seq,fusion_type,fusion['gene_name1'],fusion['gene_name2'])
  inputFile.close()

  ## Process Trinity data and compare to deFuse
  matched_transcripts = dict()
  matched_orfs = dict()
  fusions_with_transcripts = set()
  fusions_with_orfs = set()
  n = 0
  if options.transcripts: 
    with open(options.transcripts) as fp:
      for name, seq in read_fasta(fp):
        n += 1
        for i,fusion in enumerate(fusions):
          if fusion['fwd_seq'] in seq or fusion['rev_seq'] in seq:
            fusions_with_transcripts.add(i)
            matched_transcripts[name] = seq
            fusion['transcripts'].append(name)
            fusion['Transcript'] = 'Yes'
    print >> sys.stdout, "fusions_with_transcripts: %d  %s\n matched_transcripts: %d" % (len(fusions_with_transcripts),fusions_with_transcripts,len(matched_transcripts))
    for i,fusion in enumerate(fusions):
      print >> sys.stdout, "%4d\t%6s\t%s\t%s\t%s\t%s\t%s\t%s" % (i,fusion['cluster_id'],fusion['fwd_seq'],fusion['rev_seq'],fusion['fusion_type'],fusion['gene_name1'],fusion['gene_name2'], fusion['transcripts'])
    ## Process ORFs and compare to matched deFuse and Trinity data.
    ## Proteins must be at least 100 aa long, starting at the first "M" and must end with an "*".
    if options.peptides: 
      with open(options.peptides) as fp:
        for name, seq in read_fasta(fp):
          n += 1
          if len(seq) < options.min_pep_len:
            continue
          for i,fusion in enumerate(fusions):
            if len(fusion['transcripts']) > 0:
              for id_string in fusion['transcripts']:
                tx_id = id_string.lstrip('>').split()[0]
                if tx_id in name:
                  pep_len = len(seq)
                  start = seq.find('M')
                  if pep_len - start < options.min_pep_len:
                    continue
                  fusions_with_orfs.add(i)
                  matched_orfs[name] = seq
                  fusion['Protein'] = 'Yes'
                  # fwd or reverse
                  tx_seq = matched_transcripts(tx_id)
                  pos = tx_seq.find(fusion['fwd_seq'])
                  if pos < 0:
                    pos = tx_seq.find(fusion['rev_seq'])
                  # locate fusion in transcript
                  # locate fusion in ORF
                  fusion['prior_pep_seq'] = ''
                  fusion['novel_pep_seq'] = ''
      print >> sys.stdout, "fusions_with_orfs: %d  %s\n matched_orfs: %d" % (len(fusions_with_orfs),fusions_with_orfs,len(matched_orfs))
  ## Write reports
  report_fields = ['gene_name1','gene_name2','span_count','probability','gene_chromosome1','gene_location1','gene_chromosome2','gene_location2','fusion_type','Transcript','Protein']
  report_colnames = {'gene_name1':'Gene 1','gene_name2':'Gene 2','span_count':'Span cnt','probability':'Probability','gene_chromosome1':'From Chr','gene_location1':'Fusion point','gene_chromosome2':'To Chr','gene_location2':'Fusion point','fusion_type':'Type','Transcript':'Transcript?','Protein':'Protein?' }
  print >> outputFile,"%s\t%s" % ('#','\t'.join([report_colnames[x] for x in report_fields]))
  for i,fusion in enumerate(fusions): 
    print >> outputFile,"%s\t%s" % (i + 1,'\t'.join([str(fusion[x]) for x in report_fields]))
    # print >> outputFile, "%d\t%s\t%s\t%d\t%f\t%s\t%s\t%s\t%s\t%s\t%s\t%s" % (i,fusion['gene_name1'],fusion['gene_name2'],fusion['span_count'],fusion['probability'],fusion['gene_chromosome1'],fusion['gene_location1'],fusion['gene_chromosome2'],fusion['gene_location2'],fusion['fusion_type'],fusion['Transcript'],fusion['Protein'])

if __name__ == "__main__" : __main__()