Hi,

I am trying to align reads to miRbase however, I get very low alignment scores and I do not know why. I read in other entries that adapter contamination could be a problem when using bowtie(1) but I checked my fastq files and I get bona fide miRnas when I search reads the miRbase.

The library was generated with the NEBNext Small RNA library. On the website I see that it is recommended to use "AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC" for trimming, so I did

###Trimming###

#!/bin/bash

for i in $(ls *R1* | sed s/_R1.fastq.gz// | sort -u); 
do

trim_galore ${i}_R1.fastq.gz -q 20 --fastqc --cores 3 --stringency 4 --length 18 --max_length 31 --adapter AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC -o /any/directory/

done

The fastqc output looks a bit messy but I think it is expected due to the low complexity of small RNAs, nevertheless here some screenshots from a representative example:

Here is the summary outbput from trimGalore!

=== Summary ===

Total reads processed: 12,028,354 Reads with adapters:
6,119,061 (50.9%) Reads written (passing filters): 12,028,354 (100.0%)

Total basepairs processed: 613,446,054 bp Quality-trimmed:
806,268 bp (0.1%) Total written (filtered): 440,154,599 bp (71.8%)

I then checked the reads quickly and aligned some to mirBase:

@VH00411:17:AAALWCMHV:1:1101:55440:1019 1:N:0:AGTCAA TTTGGCAATGGTAGAACTCACACCG + CCCCCCCCCCCCCCCCCCCCCCCCC @VH00411:17:AAALWCMHV:1:1101:60363:1019 1:N:0:AGTCAA TTCACAGTGGCTAAGTTCTGC + ;CCCCCCCCCCCCCCCCCCCC @VH00411:17:AAALWCMHV:1:1101:45271:1038 1:N:0:AGTCAA TAGCTTATCAGACTGATGTTGA

and I get a perfect match:

I downloaded the latest miRbase (22.1) files mature.fa and I generated an index with the mature.fa file Feeling confident about the results so far, I pursue the alignment with bowtie(1)

 #!/bin/bash
basename=mature_bowtie

for i in $(ls any/directory/*.fq  | xargs -n 1 basename | sort -u); 


do

bowtie -v 2 -a --best --strata -x mature_bowtie any/directory/${i} -S any/directory/${i}.sam

done

And then I get disappointed by seeing the following output:

reads processed: 5009934

reads with at least one alignment: 62 (0.00%)

reads that failed to align: 5009872 (100.00%) Reported 213 alignments

What am I doing wrong? Do I have to convert T to U?