7
$\begingroup$

Is there a way to filter bases in BAM files based on phred quallities through python's pysam ?

I have a code here that

  • Takes the nucleobases per position from a BAM file using pysam's pileup function
  • Saves it in ReverseList and ForwardList based on both strands (i.e forward and reverse), I want to reject those bases that have phred quality below 25 to be not stored in the ForwardList and ReverseList lists so that they are not used for further analysis.

     samfile = pysam.Samfile( filename, "rb" )
     ReverseList = [''] * lenref
     ForwardList = [''] * lenref
     for pileupcolumn in samfile.pileup() :
         for pileupread in pileupcolumn.pileups:
             if (pileupread.alignment.mapping_quality <= 15):
                   continue      
             if not pileupread.is_del and not pileupread.is_refskip:
                   if pileupread.alignment.is_reverse: #negative
                          ReverseList[pileupcolumn.pos] += pileupread.alignment.query_sequence[pileupread.query_position]
                   else:
                          ForwardList[pileupcolumn.pos] += pileupread.alignment.query_sequence[pileupread.query_position]
    
     samfile.close()
    

Where lenref = 16569(length of mitochondrial genome) and filename is the name of BAM file. I want to filter based of phred qualities of bases.

$\endgroup$
4
  • 1
    $\begingroup$ Am I the only one who is missing the point of filtering alignments by phred scores? Phred scores are assigned during base calling, if you want to filter them out, why after mapping but not before? Also variant callers take into account phred scores, having low phred scores should not worsen quality of calling variants. $\endgroup$
    – Kamil S Jaron
    Jul 25 '17 at 11:46
  • $\begingroup$ I am making a tool for analyzing BAM file so I want to filter it again to improve my results. $\endgroup$ Jul 25 '17 at 12:06
  • $\begingroup$ It is also reported in previous studies. $\endgroup$ Jul 25 '17 at 12:07
  • $\begingroup$ Welcome to Bioinformatics! This question seem to need more information to get a good answer (Although the existing one answers how to filter as required). Why do you want to filter the bases? What are you trying to achieve? $\endgroup$
    – llrs
    Jul 25 '17 at 13:04
4
$\begingroup$

The PileupRead object as a query_position attribute, which you can use for this:

for pileupcolumn in samfile.pileup() :
    for pileupread in pileupcolumn.pileups:
         if (pileupread.alignment.mapping_quality <= 15):
             continue      
         if not pileupread.is_del and not pileupread.is_refskip:
             if pileupread.alignment.query_qualities[pileupread.query_position] < 10:
                 # Skip entries with base phred scores < 10
                 continue
             if pileupread.alignment.is_reverse: #negative
                   ReverseList[pileupcolumn.pos] += pileupread.alignment.query_sequence[pileupread.query_position]
             else:
                   ForwardList[pileupcolumn.pos] += pileupread.alignment.query_sequence[pileupread.query_position]

Note the 6th line, which implements a filter with a threshold of 10.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.