SEQanswers

Go Back   SEQanswers > Applications Forums > Sample Prep / Library Generation



Similar Threads
Thread Thread Starter Forum Replies Last Post
Twin peaks, double peaks, double humps of RNA Seq, it’s all very frustrating...HELP! AndyG Sample Prep / Library Generation 18 10-24-2015 04:43 PM
Latest Edition : On the trail of Genomic Pioneers interview series manoj.b Literature Watch 1 04-12-2010 05:55 AM
The inaugural edition of our ‘On the Trail of Genomic Pioneers’ interview series manoj.b General 0 10-14-2009 09:05 AM

Reply
 
Thread Tools
Old 02-10-2015, 08:42 AM   #1
franz263172
Junior Member
 
Location: Vienna, Austria

Join Date: Feb 2015
Posts: 3
Default Twin Peaks - miRNA Edition

Hello everyone,

My name is Franz, I'm new to seqanswers.com, and also pretty new to NGS in general, so I hope my questions aren't too stupid.

Well, I want to do miRNA sequencing on Ion Torrent platform, using their Ion Total RNA seq kit v2.
I seem to have gotten it to a point where it seems to work fine, except this huge peak just below the miRNA inserts keeps popping up all the time. I know its not adaptor dimers, those should be smaller. In the case of the Ion Torrent kit, the adaptor dimers should be ~77nt, and the miRNAs ~97nt. I can get rid of the dimers using the included magnetic beads, and I get the miRNA peak, however, the additional peak I'm having is at ~87nt (insert-length ~10).

I do not have smallRNA chips for the bioanalyzer, so I can't say if it isn't simply actual RNA inserts (would degradation fragments actually look like this?), and it also might be the case that the input amount is not right. I am extracting the RNA from fresh-frozen human tissue using the Invitrogen PureLink miRNA kit. The input amount should be 1-100ng miRNA, with 5-25ng being ideal, I varied the input amount from 60ng smallRNA to 300ng, perhaps its not enough?
Also, I put some libraries on the bioanalyzer right after the final amplification step before size-selectioning, and it seems that the beads work for the dimers and some of the tRNA inserts above, but not for that 87nt peak, is it simply too close?

I have attached a PDF with bioanalyzer traces. I know the majority of people don't use Ion Torrent, but I believe the other manufacturer's kits work similarly. Does anyone here have any idea what this might be? A peak right between the adaptor dimers and the miRNA inserts?



regards,
Franz
Attached Files
File Type: pdf smallRNA libraries.pdf (1.22 MB, 35 views)

Last edited by franz263172; 02-10-2015 at 08:57 AM. Reason: forgot attachment
franz263172 is offline   Reply With Quote
Old 02-10-2015, 04:03 PM   #2
cascoamarillo
Senior Member
 
Location: MA

Join Date: Oct 2010
Posts: 158
Default

...Interesting.
Not sure how Invitrogen PureLink miRNA kit works with size selecting small RNAs or Ion Total RNA seq kit do the libraries. Anyway, some questions/ideas. Fresh-frozen tissue means fresh tissue was frozen and afterwards RNA extracted or it is been long time frozen?
The proportion seems to be 50:50. I'd recommend clone it and sequence like 100 clones and see the real size and what's in there.
Are all the libraries equal in final concentration? If so, I don't think the 77 peak should be a real threat (don't know too much about ion torrent sequencing bias...). The best way to get rid of it: PAGE denaturing gel. Run your samples in like 15% PAGE and size select your libraries; you won't get a single dimer in the sequence. Also would be interesting look at the double peak in these conditions. If the clone analysis shows 89 peak is crap, you can try to get rid of it too.
Good luck
cascoamarillo is offline   Reply With Quote
Old 02-11-2015, 12:43 AM   #3
franz263172
Junior Member
 
Location: Vienna, Austria

Join Date: Feb 2015
Posts: 3
Default

it was long time frozen tissue.
I should mention, I once sequenced a library that had the 10-insert peak, and judging from the length distribution of the reads it did sequence, however, I wasn't able to find out what it is, probably due to my lack of bioinformatics knowledge.
how would i go about finding out if the peak is clonal?
franz263172 is offline   Reply With Quote
Old 02-11-2015, 09:44 AM   #4
cascoamarillo
Senior Member
 
Location: MA

Join Date: Oct 2010
Posts: 158
Default

Quote:
Originally Posted by franz263172 View Post
..., and judging from the length distribution of the reads it did sequence, however, I wasn't able to find out what it is, probably due to my lack of bioinformatics knowledge.
how would i go about finding out if the peak is clonal?
Sorry, I didn't get that. So you did sequence it and a 89bp construct (10 nt insert) showed up? Well, were all of them the same sequence?
There are plenty of programs for miRNA analysis, specially in model organism (in this case human). You can check a few in the SEQanswers wiki:
http://seqanswers.com/wiki/Software/list
There are a few ones which are online, like miRanalyzer:
http://bioinfo5.ugr.es/miRanalyzer/miRanalyzer.php
cascoamarillo is offline   Reply With Quote
Old 02-11-2015, 02:19 PM   #5
franz263172
Junior Member
 
Location: Vienna, Austria

Join Date: Feb 2015
Posts: 3
Default

yes, pretty much. although i wasn't very precise when i said 10nt. its a peak around 10nt, not that broad, but not all of its reads are exactly 10nt. in fact it looks just like in the bioanalyzer curve in the read length histogram.

i think it might be degradation fragments after all.
i uploaded the fastq file to a galaxy server, removed adaptors and low quality reads according to ion torrents manual, and then filtered only the reads between 10-14nt, which i mapped to a fasta file with just the human rRNAs using bowtie2. well, 60% of it mapped.
does that make sense, or do you have too much ambiguity/chance mapping a 10-14nt read even with a small reference? (this is the file i used)
franz263172 is offline   Reply With Quote
Old 02-12-2015, 01:26 PM   #6
cascoamarillo
Senior Member
 
Location: MA

Join Date: Oct 2010
Posts: 158
Default

Better try to use Bowtie1 for small fragment analysis. Algorithms apart, it seems you have rRNA contamination into your miRNA library. Well, this happens sometimes. The good thing is you still can get a 40% (of don't know how many million reads from IonTorrent) with potential candidates. Still, if you are looking for miRNA try to compare them with known miRNA databases; there are many programs for that, taking into account conservative regions, etc etc, which can give you a better idea.
cascoamarillo is offline   Reply With Quote
Old 02-13-2015, 12:47 AM   #7
sarvidsson
Senior Member
 
Location: Berlin, Germany

Join Date: Jan 2015
Posts: 137
Default

Quote:
Originally Posted by franz263172 View Post
yes, pretty much. although i wasn't very precise when i said 10nt. its a peak around 10nt, not that broad, but not all of its reads are exactly 10nt. in fact it looks just like in the bioanalyzer curve in the read length histogram.

i think it might be degradation fragments after all.
i uploaded the fastq file to a galaxy server, removed adaptors and low quality reads according to ion torrents manual, and then filtered only the reads between 10-14nt, which i mapped to a fasta file with just the human rRNAs using bowtie2. well, 60% of it mapped.
does that make sense, or do you have too much ambiguity/chance mapping a 10-14nt read even with a small reference? (this is the file i used)
I've also seen degraded miRNA in that size range coming through sequencing along with the full length versions. You can check alignment accuracy by separately mapping the two groups of reads and comparing the relative abundances in the end.

However, I would rather filter out the degradation products and just keep the good stuff - typically you don't need that deep sequencing to cover miRNA diversity anyway...
sarvidsson is offline   Reply With Quote
Reply

Tags
bioanalyzer, ion torrent, library preparation, mirna, mirna seq problem

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 05:57 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2018, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO