Low mapping quality from LifeScope alignment

[endether]

Dear All,

I am working on a pair-end (75+35bp) SOLiD dataset. We used LifeScope to perform the analysis (mapping to genome and reads counting on genes etc.). However, we found that the mapping quality output by LifeScope is really low.

For example, in one library, we have around 66 million(M) raw reads. 10M of them were unmapped. Among mapped reads, 45M reads have an alignment score < 10 (most of them are actually 0).

We double checked our experiment settings, and sequencing machine also reported good quality on raw reads. We are now confused about which part can be wrong.

I looked at those low quality alignment from the BAM file, reads like following record has mapping quality 0, which is anti-intuitive to us:

698_1442_2018 113 Chr01 200535 0 3S72M = 7706595 7506094 TGAGATGATTAACATATAAATCTGTAGCTACATGGAATTAAGGTAGTGGAGCAGAGGAGGAAGGATGATAGAAGA 6;@JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ RG:Z:1_1 NH:i:2 CM:i:0 NM:i:0 CQ:Z:@@@@@@@@@@@@@@>@@@@@@@@@@@@@@@@@@@@@@@@?@;@6@@@@?@@?=@@?6?@@@@=<@@@@@@@@@<8 CS:Z:T022022332132020202202221322011231020303020131132323112230033331103032132221



Could anyone shed some light on what might go wrong?

Thanks so much!

Best,
Zheng

[TiborNagy]

Zero mapping quality usually means the reads can be mapped to multiple locations. You can check this if you enable multiple mappings on you aligner program or cut one of these regions and run a Blast search with it.

[endether]

Zero mapping quality usually means the reads can be mapped to multiple locations. You can check this if you enable multiple mappings on you aligner program or cut one of these regions and run a Blast search with it.

Thank you so much for the reply. I cannot find such an option in LifeScope allowing multiple mappings. I think you are right, we took reads with 0 quality alignment, and realign them in code space in Bowtie, and indeed lots of them have more than one hit.

Thanks you again for the help.

Best,
Zheng