This is a test version of Biostars. For the public version, visit https://www.biostars.org.
smallRNA alignment and counting multimapping

Hello! I am analyzing smallRNA data of human genome. I have aligned my read against genome with bowtie2:

bowtie2 --end-to-end --very-sensitive -x hg38 -U input_file -S output_file_align

and I used featurescount in order to obtained the counts. For the counting: I used the gtf file with miRNA: featureCounts -T 5 -t exon -g gene_name -a GTF_miRNA -R SAM -o fileOutput output_file_align and the reads not assigned to miRNA were annotated with the gtf from gencode for the other non-coding RNA molecules.

This is the result of the alignment:

 41798971 reads; of these:
  41798971 (100.00%) were unpaired; of these:
    1101846 (2.64%) aligned 0 times
    10724437 (25.66%) aligned exactly 1 time
    29972688 (71.71%) aligned >1 times
97.36% overall alignment rate

This is the results of featurescount for miRNA:

Total reads : 41798971                                                 
Successfully assigned reads : 15157113 (36.3%)

And this for the other non-coding RNA molecules:

Total reads : 41798971                                                 
 Successfully assigned reads : 12408845 (29.7%)

As reported in the featurescount manual:

For the reporting of a multi-mapping read, users may choose to not report any alignments for the read (by default) or report up to a pre-defined number of alignments (‘–multiMapping’ and ‘-B’ options).

How is it possible that I obtained the previously result being the 71.71% of read aligned >1 times?

Thank you in advance

smallrna alignment featurescount

1 answer

reads not assigned to miRNA were annotated

I don't think that's right since the number of total reads is the same in both featureCount cases

In regard to featureCounts not recognizing the multi-mappers -- I'm not sure how what kind of information featureCounts is relying on (re-check the manual, maybe it mentions it). It may need the flag that is typically set by Picards MarkDuplicates, which bowtie2 probably does not do (if I recall correctly, bowtie2 reports multi-mappers via optional tags in the SAM file (e.g. XS), not via the flag in the fifth column of the SAM file).

Log in to answer this question.