Hi
I'm mapping single-end illumina reads to a reference transcriptome assembled de-novo with GS-Assembler (Newbler) from 454 generated data, as there is no reference genome for the organism in question (a basidiomycete fungus). I'm using bowtie2 to do this and it works quite well, as I'm able to map more than 80% of my reads:
Bowtie2 stats:
24570701 reads; of these:
24570701 (100.00%) were unpaired; of these:
4447149 (18.10%) aligned 0 times
8014152 (32.62%) aligned exactly 1 time
12109400 (49.28%) aligned >1 times
I've extracted the raw read counts with samtools idxstats, and summed the reads for isoforms of the same gene (isogroup) in order to test for differential gene expression with DESeq or EdgeR.
In my reference transcriptome, which consists of isoforms (isotigs) I expect many reads to map more than one place and I have therefore used the -M flag at 100 to ensure that up to 100 relevant mapping hits are considered when running bowtie2. However, I'm wondering how bowtie2 assigns reads that map with the same quality and alignment score to more than one place?
Ideally I would like to have these reads assigned randomly between equally best hit positions so my summation for each gene is valid, but I haven't been able to find this information in the manual, paper or supplement to the paper. I would be very grateful for any comments or help on this.
I'm mapping single-end illumina reads to a reference transcriptome assembled de-novo with GS-Assembler (Newbler) from 454 generated data, as there is no reference genome for the organism in question (a basidiomycete fungus). I'm using bowtie2 to do this and it works quite well, as I'm able to map more than 80% of my reads:
Bowtie2 stats:
24570701 reads; of these:
24570701 (100.00%) were unpaired; of these:
4447149 (18.10%) aligned 0 times
8014152 (32.62%) aligned exactly 1 time
12109400 (49.28%) aligned >1 times
I've extracted the raw read counts with samtools idxstats, and summed the reads for isoforms of the same gene (isogroup) in order to test for differential gene expression with DESeq or EdgeR.
In my reference transcriptome, which consists of isoforms (isotigs) I expect many reads to map more than one place and I have therefore used the -M flag at 100 to ensure that up to 100 relevant mapping hits are considered when running bowtie2. However, I'm wondering how bowtie2 assigns reads that map with the same quality and alignment score to more than one place?
Ideally I would like to have these reads assigned randomly between equally best hit positions so my summation for each gene is valid, but I haven't been able to find this information in the manual, paper or supplement to the paper. I would be very grateful for any comments or help on this.