Error in bowtie's sam output?

[burt]

Hi,

I'm not sure if its a bug with the bowtie's sam output, but all my mapping quality would take the value of 255 while the bit flag would take on values of either 0,4,16. I had run bowtie in colorspace.

Has anyone else encountered a similar issue?

Here's a sample of the sam output with the qwerky values.

1279_33_430_F3 4 * 0 0 * * 0 0 TGGTGACCTGGGCCCTGAGGANCGCGTTGATCTACCTCCGCTCAATTGT IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII XM:i:0
1279_33_409_F3 16 chr6 49186545 255 50M * 0 0 TATTCGTACTGAAAATCAAGATCAAGCGAGCTTTTGCCCTTCTGCTCCAC Iqqqqqqqqqqqqqqqqqqqqqqqqqq!Iq!!qqqqqqqqqqqqqqqqqI XA:i:2 MD:Z:50 NM:i:0 CM:i:2
1279_34_783_F3 0 chr16 11144009 255 50M * 0 0 TATGTGCTTGGCTGAGGAGCCAATGGGGCGAAGCTACCATCTGTGGGATT Iqqqqqqqqqqqqqqqqqqqq!Iqqqqqqqqqqqqqqqqqqqqqqq!!qI XA:i:2 MD:Z:50 NM:i:0 CM:i:2
1279_40_121_F3 0 chr17 39983712 255 50M * 0 0 ACGGGGAATCAGGGTTCGATTCCGGAGAGGGAGCCTGAGAAACGGCTACC Iqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqq!!qqq!!qqqqI XA:i:2 MD:Z:50 NM:i:0 CM:i:2
1279_41_567_F3 16 chr6 49186546 255 50M * 0 0 ATTCGTACTGAAAATCAAGATCAAGCGAGCTTTTGCCCTTCTGCTCCACG IqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqI XA:i:0 MD:Z:50 NM:i:0 CM:i:0


-burt

[me_myself_andI]

Related and up to now also unanswered post: http://seqanswers.com/forums/showthread.php?t=10624

[Joker!sAce]

Field MAPQ considers pairing in calculation if the read is paired. If such a calculation is difficult, 255 is applied, indicating the mapping quality is not available. I assume your query sequences had quality scores with them(FastQ files). Maybe your query sequence was not paired? There is a high order of probability that this is not an error and you should not worry about it, but confirm this.

[burt]

Hi, thanks for the reply. My input data is a set of single-end reads. I'm pretty sure if a problem with the sam output in colorspace.

Just something to add on to the problem description. If I were to look only are reads that are mappable, all of them are reported to have 50M perfect matching. This is very unusual as it's highly unlikely for all reads to map perfectly to the reference genome, especially when we are investigating the transcriptome landscape of the cell.

[feixue1039]

Hi

I encountered the same issue as burt. My input data is single-end reads sequenced by Illumina Hiseq 2000 platform. I just wonder whether a read with a MAPQ value of 255 is filtered or not before the calculation of FPKM value in downstream analyses (for instance, cufflinks/cuffdiff)?

Any reply will be appreciated.

feixue1039

[swbarnes2]

If I were to look only are reads that are mappable, all of them are reported to have 50M perfect matching.

That's not what the CIGAR score means. 50M only means no indels, no hard or soft clipping at the ends. There are other elements later in the sam line that indicate what the discrepancies exist between the read and the reference.

[loodramon]

Hi,

I am seeing the same thing with my single stranded (non-paired) RNA-seq alignments.

There is either a Mapping Quality score of 255 and has a bit flag of 16 or else the read is not mapped and has a bit flag of 4.

Is this common to all non-paired RNA-seq data? I'm new to RNA-seq.

Thanks in advance

[loodramon]

Hi again,

I should mention that I used the latest version of bowtie for this alignment.

I've been told by a colleague that: Bowtie doesn't calculate mapping quality values, so it prints 255 to the MAPQ field of the sam file if the read aligns or 0 otherwise.

[edge]

Hi,

Do you know what is the meaning of 100M from bowtie sam output?
I run bowtie with single-read.
My input file is 100 read length.

Thanks.

[edge]

Hi.

Thanks for answer about 255 by bowtie sam output.

Hi,

Do you know what is the meaning of 100M from bowtie sam output?
I run bowtie with single-read.
My input file is 100 read length.

Thanks.

[dpryan]

Hi,

Do you know what is the meaning of 100M from bowtie sam output?
I run bowtie with single-read.
My input file is 100 read length.

Thanks.

It's generally better to start a new thread rather than to resurrect a really old one. Nevertheless, the "100M" is part of the CIGAR string, which is defined in the SAM specification. It means 100 matches (practically, this just means that there were no indels). I recommend familiarising yourself with the SAM format if you're going to do much with sequencing data.

[edge]

Thanks, dpryan.

I will have a look on the document that you shared.
Really appreciate