yhwu / sam2psl

Convert SAM format to PSL format

sam2psl

Convert SAM format to PSL format.

Usage

   make  
   ./sam2psl -h  ## print header, control-c to exit  
   cat tmp.sam | ./sam2psl -h  | grep -v ^#

Notes

• This software is written to be portable, meant to be used in a pipe. To compile, g++ -O2 sam2psl.cpp. Bug reports are apprecaited.
• By default, the original SAM alignments are printed with '#' at the beginning. Use grep -v ^# to get rid of them. Use | cut -f-21 to produce PSL only output.
• This software is only tested on bwa and bowtie2 outputs and may or maybe not work well for other aligners.
• tStart, is reported as -1 if a read is not aligned.
• matches, is the total length of M/I/D blocks from CIGAR string.
• misMatches, is the edit distance reported in the NM:i:[0-9]+ field.
• repMatches, is not calculated, reported as 0.
• blocks, in SAM format, all M/I/D/= blocks are considered matched parts. To closely conform with PSL format, all M blocks are treated as different blocks. For example, a CIGAR with 10S30M4D30M5S produces 2 blocks, with blockSizes being 30,30,, qStarts being 10,40,, and tStarts being POS-1, POS-1+34,.
• TLEN, is the whole template length reported by bowtie2, matched(not including soft-cliped part) template length reported by BWA.
• AS, is the mapping score. Different aligners have different formulars.
• MAPS, is given by matches-misMatches; this value is calculated so that alignments from different aligners are comparable.

Status

• qBlocks and tBlocks are not printed yet, but the two fields are present in the output as ,'s.
• The other fields should be accurate.

Output

The columns are:

matches,     ##1.  Number of matching bases that aren't repeats; note, all matches are included. 
misMatches,  ##2.  Number of bases that don't match.
repMatches,  ##3.  Number of matching bases that are part of repeats; always 0, see column 1.  
nCount,      ##4.  Number of 'N' bases.
qNumInsert,  ##5.  Number of inserts in query.
qBaseInsert, ##6.  Number of bases inserted into query.
tNumInsert,  ##7.  Number of inserts in target.
tBaseInsert, ##8.  Number of bases inserted into target; commonly refered as deletion.
strand,      ##9.  '+' or '-'; '*' if cannot be determined.
qName,       ##10. Query sequence name.
qSize,       ##11. Query sequence size.
qStart,      ##12. Alignment start position in query.
qEnd,        ##13. Alignment end position in query.
tName,       ##14. Target sequence name.
tSize,       ##15. Target sequence size.
tStart,      ##16. Alignment start position in query.
tEnd,        ##17. Alignment end position in query.
blockCount,  ##18. Number of blocks in the alignment.
blockSizes,  ##19. Comma-separated list of sizes of each block.
qStarts,     ##20. Comma-separated list of start position of each block in query.
tStarts,     ##21. Comma-separated list of start position of each block in target.

qBlocks,     ##22. Comma-separated list of sequence blocks in query. 
tBlocks,     ##23. Comma-separated list of sequence blocks in target on aligned strand reading from 5' to 3'. 

RI,          ##24. Read index, [1/2].
RNEXT,       ##25. Mate target, [=/chr].
PNEXT,       ##26. Mate position, 0-based.
TLEN,        ##27. Template length, reported by aligner. 
MAPQ,        ##28. Mapping quality.
AS,          ##29. Mapping score, reported by the aligner.
MAPS,        ##30. Mapping score, matched length - edit distance, similar to bwa.
FPAIRED,     ##31. Paired end or single end, [P/S].
FPROPER_PAIR,##32. Pairs properly paired or unpaired, [P/U].
FSECONDARY,  ##33. Primary or secondary mapping, [P/S].    
FQC,         ##34. Quality control pass or fail, [P/F].
FDUP,        ##35. Read primary or duplicated, [P/D].