rRNA reads classified as others

[Kec89]

Hi Andrew,

I used the human genomic fasta and gtf annotation you suggested, but I am having an issue regarding the reads classified as "others". In my case I know they are rRNA reads, and I would like them to be classified as such. I thought the issue was related to where on the genome rRNAs are annotated, so I tried different fasta/gtf combinations but nothing worked so far. Do you have any suggestion? This problem is particularly important as reads classified as "others" do not seem to contribute to the normalization factors DeSeq2 estimates. Do you know how can I make them contribute to DeSeq2?

Thanks for your great work and help! Tea

[patriciaplchan]

Hi Tea,

The GTF file defines the non-tRNA transcripts to be included in the analysis. For your case, it seems like the GTF file does not include the annotations of those rRNAs that you are interested. You may want to add those annotations manually in the GTF file and rerun the analysis. tRAX uses ENSEMBLE formatted GTF file. So long as your file follows the same format, tRAX should work accordingly.

best, Patricia

[andrewdholmes]

If you have a bed file or a set of bed files with these features you can use the --bedfiles option to include those features in addition to the GTF file features. Each bed file will count as it's own gene type in downstream analyses

On Tue, Apr 6, 2021 at 7:15 AM Kec89 @.***> wrote:

Hi Andrew,

I used the human genomic fasta and gtf annotation you suggested, but I am having an issue regarding the reads classified as "others". In my case I know they are rRNA reads, and I would like them to be classified as such. I thought the issue was related to where on the genome rRNAs are annotated, so I tried different fasta/gtf combinations but nothing worked so far. Do you have any suggestion? This problem is particularly important as reads classified as "others" do not seem to contribute to the normalization factors DeSeq2 estimates. Do you know how can I make them contribute to DeSeq2?

Thanks for your great work and help! Tea

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/UCSC-LoweLab/tRAX/issues/16, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEHEMLTU7SKO5BBTKVQDFX3THMJRLANCNFSM42O2PYFA .

[Kec89]

Thanks Patricia and Andrew!

Il giorno mar 6 apr 2021 alle ore 22:52 andrewdholmes < @.***> ha scritto:

If you have a bed file or a set of bed files with these features you can use the --bedfiles option to include those features in addition to the GTF file features. Each bed file will count as it's own gene type in downstream analyses

On Tue, Apr 6, 2021 at 7:15 AM Kec89 @.***> wrote:

Hi Andrew,

I used the human genomic fasta and gtf annotation you suggested, but I am having an issue regarding the reads classified as "others". In my case I know they are rRNA reads, and I would like them to be classified as such. I thought the issue was related to where on the genome rRNAs are annotated, so I tried different fasta/gtf combinations but nothing worked so far. Do you have any suggestion? This problem is particularly important as reads classified as "others" do not seem to contribute to the normalization factors DeSeq2 estimates. Do you know how can I make them contribute to DeSeq2?

Thanks for your great work and help! Tea

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/UCSC-LoweLab/tRAX/issues/16, or unsubscribe < https://github.com/notifications/unsubscribe-auth/AEHEMLTU7SKO5BBTKVQDFX3THMJRLANCNFSM42O2PYFA

.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/UCSC-LoweLab/tRAX/issues/16#issuecomment-814432673, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKGC7MFBS6B52GM7UZOISRLTHNYAFANCNFSM42O2PYFA .