Dear all,
We are using BWA (0.5.7 ) to align Illumina read to refseq database, however we encountered some strange results see following:
This is from a SE lane mapped against our hg19 custom transcript database. The result from the SAM file for this read is as follows (a few line breaks are added for viewing purpose):
HWI-EAS275:2:43:496:1741#0 4 SERPINA3|12|BC003559.1|13097704|mRNA|cds:53..1324 1603 0 36M * 0 0 CGGCTTTGGAATCCACCAGCTACATCCAGCTCCCTG abbbbaba`]\aba``a``aaa_]_a`]]``aa_ZB XT:A:R NM:i:0 X0:i:4 X1:i:0 XM:i:0 XO:i:0 XG:i:0 MD:Z:36 XA:Z:SERPINA3|12|BC070265.1|47124537|mRNA|cds:63..350,+13,36M,0;SERPINA3|12|NM_001085.4|73858562|mRNA|cds:80..1351,+30,36M,0;SERPINA3|12|BC003559.1|13097704|mRNA|cds:53..1324,+3,36M,0;
The odd thing about this particular example is that the SAM file reports that the match begins at position 1603 of the transcript -- but this transcript is exactly 1603 bases long, so how could there be a 36M match after this position? I realize that the 0x0004 flag indicates that the read is unmapped -- so why is this result returned at all? Why isn't it just reported like other unmapped reads, in the format of: "<readID> 4 * 0 0 * * 0 0 <seq> <qual>" ?
We have noticed PE results where an unmapped read is reported to have its mate's mapping but also the 0x0004 flag -- but this is not a PE read, it is SE.
Anyone got any idea?
Jianying Li
We are using BWA (0.5.7 ) to align Illumina read to refseq database, however we encountered some strange results see following:
This is from a SE lane mapped against our hg19 custom transcript database. The result from the SAM file for this read is as follows (a few line breaks are added for viewing purpose):
HWI-EAS275:2:43:496:1741#0 4 SERPINA3|12|BC003559.1|13097704|mRNA|cds:53..1324 1603 0 36M * 0 0 CGGCTTTGGAATCCACCAGCTACATCCAGCTCCCTG abbbbaba`]\aba``a``aaa_]_a`]]``aa_ZB XT:A:R NM:i:0 X0:i:4 X1:i:0 XM:i:0 XO:i:0 XG:i:0 MD:Z:36 XA:Z:SERPINA3|12|BC070265.1|47124537|mRNA|cds:63..350,+13,36M,0;SERPINA3|12|NM_001085.4|73858562|mRNA|cds:80..1351,+30,36M,0;SERPINA3|12|BC003559.1|13097704|mRNA|cds:53..1324,+3,36M,0;
The odd thing about this particular example is that the SAM file reports that the match begins at position 1603 of the transcript -- but this transcript is exactly 1603 bases long, so how could there be a 36M match after this position? I realize that the 0x0004 flag indicates that the read is unmapped -- so why is this result returned at all? Why isn't it just reported like other unmapped reads, in the format of: "<readID> 4 * 0 0 * * 0 0 <seq> <qual>" ?
We have noticed PE results where an unmapped read is reported to have its mate's mapping but also the 0x0004 flag -- but this is not a PE read, it is SE.
Anyone got any idea?
Jianying Li
