Re: [Transdecoder-users] Does TranscriptDecoder also predict polyA?
Extracting likely coding regions from transcript sequences
Brought to you by:
bhaas
Menu
▾
▴
Re: [Transdecoder-users] Does TranscriptDecoder also predict polyA?
|
From: Trakhtenberg, F. <Eph...@ch...> - 2014-07-27 18:25:44
|
Hi Brian, Thank you again for the feedback. If Trinity can reconstruct transcripts with ployA, I may need to learn how to use it. On Wed I posted a question on Tophat/Cufflinks board regarding this issue, but so far no responses: https://groups.google.com/forum/#!topic/tuxedo-tools-users/Ntv98xmHT7c. The PASA sounds like an excellent resource for this as well, thank you. Bets regards, Ephraim ________________________________ From: Brian Haas [bh...@br...] Sent: Tuesday, July 22, 2014 9:20 PM To: Trakhtenberg, Feliks Cc: tra...@li...; rc...@li... Subject: Re: [Transdecoder-users] Does TranscriptDecoder also predict polyA? It's definitely an interesting research question. We've noticed anecdotally that many of our Trinity-reconstructed transcripts appear to have a stretch of polyA at their 3' end (or in non-strand-specific RNA-Seq assembly, a poly T at 5' end corresponding to the rev-comp). It would be useful to know how many polyA's we can detect, from the reads and from our rna-seq assemblies. Note, we have a polyA-detection system that we used in PASA (http://pasa.sf.net), and published a couple of papers on several years ago, based on EST and cDNA data, but I haven't revisited it with more modern data types. It's certainly an area that deserves a lot more attention. best, ~b On Tue, Jul 22, 2014 at 9:01 PM, Trakhtenberg, Feliks <Eph...@ch...<mailto:Eph...@ch...>> wrote: This may actually be interesting to investigate: what percent of the transcripts with TranscriptDecoder predicted ORF have polyA site predicted in appropriate location, and how this correlates with actual polyA addition. We have polyA-selected RNAseq and total RNAseq (after ribo depletion), which we could use for addressing this question. Still trying to figure out if we could also detect actual polyA sequences on 3' RNA fragments in the raw reads. I am also interested in polyA-lncRNAs, which have no ORF but are exported to the cytoplasm because of polyA. If these questions might be of interest to you too, or if this has been done already, please let us know. thank you, Ephraim ________________________________ From: Brian Haas [bh...@br...<mailto:bh...@br...>] Sent: Tuesday, July 22, 2014 1:17 PM To: Trakhtenberg, Feliks Cc: tra...@li...<mailto:tra...@li...>; rc...@li...<mailto:rc...@li...> Subject: Re: [Transdecoder-users] Does TranscriptDecoder also predict polyA? Thanks for the info! I've been thinking about this some more, and I'm not sure how useful it will be in the context of transdecoder after all... This is because, although these tools can be quite useful for identifying or predicting sites of polyadenylation, not finding such a site might not be a robust indicator that you don't have a coding transcript, particularly if you find a nice long ORF with good coding potential. For now, I think we'll rely on coding metrics and homology data, and eventually we'll leverage a lincRNA and other ncRNA classification tools to facilitate transcriptome annotations. best, ~brian On Tue, Jul 22, 2014 at 12:30 PM, Trakhtenberg, Feliks <Eph...@ch...<mailto:Eph...@ch...>> wrote: Hi Brian, CC'ed: Lingsheng Dong from our bioinformatics support team. The 3 links below seem to offer free software for mouse polyA prediction using transcripts sequences input, which could be taken directly from Cufflinks de novo transcripts output. I wonder if one of them could be a good fit for using along with the TranscriptDecoder. For example, the polyA site that occurs on 3' past the predicted stop codon would be valid, but if there is no polyA site, or more than 10 bp before the stop codon, such transcript would probably not produce a protein even if ORF is predicted. The 3' ends are cleaved off about 10 bp downstream of the polyA site prior to polyA addition, that is why I thought polyA site could not be more than 10 bp upstream of the stop codon. Please let me know what you think about this. http://exon.umdnj.edu/polya_svm_server/index.html http://mlkd.csd.auth.gr/PolyA/datasets.html http://genes.mit.edu/GENSCAN.html Thank you, Ephraim P.S. Links below seem to be either only for Human or predict polyA site directly from genome DNA rather than from spliced transcripts sequences: http://grail.lsd.ornl.gov/grailexp/ http://linux1.softberry.com/berry.phtml?topic=products http://www.imtech.res.in/raghava/polyapred/help.html http://cbrc.kaust.edu.sa/dps/ http://dnafsminer.bic.nus.edu.sg/PolyA.html http://cub.comsats.edu.pk/polyapredict.htm ________________________________ From: Brian Haas [bh...@br...<mailto:bh...@br...>] Sent: Saturday, July 19, 2014 6:33 PM To: Trakhtenberg, Feliks Cc: tra...@li...<mailto:tra...@li...> Subject: Re: [Transdecoder-users] Does TranscriptDecoder also predict polyA? I actually haven't researched that in a while. Please let me know if you find something really useful. best, ~brian On Sat, Jul 19, 2014 at 6:00 PM, Trakhtenberg, Feliks <Eph...@ch...<mailto:Eph...@ch...>> wrote: Thank you for the clarification. Any recommendations on which tool I could use for predicting polyA? ________________________________ From: Brian Haas [bh...@br...<mailto:bh...@br...>] Sent: Saturday, July 19, 2014 5:53 PM To: Trakhtenberg, Feliks Cc: tra...@li...<mailto:tra...@li...> Subject: Re: [Transdecoder-users] Does TranscriptDecoder also predict polyA? Hi, TransDecoder does not take into account whether a transcript is polyadenylated or not. It strictly looks at open reading frames and coding potential. best, ~brian On Sat, Jul 19, 2014 at 2:38 PM, Trakhtenberg, Feliks <Eph...@ch...<mailto:Eph...@ch...>> wrote: Hello, I am new to TranscriptDecoder. Does it consider whether a novel transcript with a predicted ORF is also predicted to be polyadenylated? I assumed if no polyA can be predicted, then even if ORF is predicted, such transcript may not produce a protein. My input is a group of novel transcripts predicted by Cufflinks analysis. Would appreciate feedback. thank you ------------------------------------------------------------------------------ Want fast and easy access to all the code in your enterprise? Index and search up to 200,000 lines of code with a free copy of Black Duck Code Sight - the same software that powers the world's largest code search on Ohloh, the Black Duck Open Hub! Try it now. http://p.sf.net/sfu/bds _______________________________________________ Transdecoder-users mailing list Tra...@li...<mailto:Tra...@li...> https://lists.sourceforge.net/lists/listinfo/transdecoder-users -- -- Brian J. Haas The Broad Institute http://broad.mit.edu/~bhaas -- -- Brian J. Haas The Broad Institute http://broad.mit.edu/~bhaas -- -- Brian J. Haas The Broad Institute http://broad.mit.edu/~bhaas -- -- Brian J. Haas The Broad Institute http://broad.mit.edu/~bhaas |
