mapping rna-seq to bacterial protein coding genes

[Francy87]

Hello!
I want to map my metatranscriptome to a set of protein coding genes from the sequenced genomes of the species usually found in this environment. I downloaded the .ffn files of the species of interest from ncbi and then I clustered them at 95% of identity with uclust. Then I tried to map my reads using bowtie2 in local mode (--sensitive-local). I got an average of 15-30% of reads aligned. Which parameters could I modify to increase the alignment rate? My reads are single ends (150bp, before adapters trimming). I'm a bit surprised because I know (from the 6S data) that the species I used to build my index are the 90% of the microbiota...

Any suggestion?

[SylvainL]

Hi, are you sure the rRNA depletion was efficient?

[Francy87]

well, I'm not 100% sure, but it should be. I used the ribo zero kit, that should be the best. How could I check for this?

[SylvainL]

The fastest would be to run a fastQC to check for over-represented sequences... The best would be to map on the full genomes (really long, depending on how many species of interest you have...)