SEQanswers

Go Back   SEQanswers > Sequencing Technologies/Companies > SOLiD



Similar Threads
Thread Thread Starter Forum Replies Last Post
SOLiD unmapped RNA-seq reads from TOPHAT rkk SOLiD 1 01-18-2012 04:18 AM
SOLiD unmapped RNA-seq reads from TOPHAT rkk Bioinformatics 0 12-22-2011 01:40 PM
Problem using TopHat with SOLiD SE-reads Boofer Bioinformatics 0 01-19-2011 03:57 AM
solid reads stuck at searching juntions in TopHat zorph Bioinformatics 8 12-13-2010 04:50 PM
tuning the tophat/cufflink pipeline oliviera RNA Sequencing 2 06-08-2010 11:49 PM

Reply
 
Thread Tools
Old 10-15-2010, 07:46 AM   #1
Pejman
Member
 
Location: Switzerland

Join Date: Jul 2010
Posts: 23
Default Tuning TopHat parameters for SOLiD reads

Hi folks

The new TopHat 1.1.1, nicely handles SOLiD reads, and the results can be piped easily to Cufflinks to make up a RNAseq pipeline. However, the manual page says:

Quote:
In TopHat 1.1.0, we began supporting Applied Biosystems' Colorspace format. The software is optimized for reads 75bp or longer.
and there are no further guidelines how to tune the parameters if you have shorted reads, as in my case 30bp single-end. Does anybody have any clue on this? Of course this is tunable through editing the TopHat scripts but, the question is how to set the parameters not to screw the whole thing up

I did a comparison, using TopHat+GTF file and defaults parameters, I get about 5 times less aligned tags than what I get from Bowtie with some sensible parameters.

Last edited by Pejman; 10-15-2010 at 08:59 AM.
Pejman is offline   Reply With Quote
Old 10-15-2010, 12:31 PM   #2
mrawlins
Member
 
Location: Retirement - Not working with bioinformatics anymore.

Join Date: Apr 2010
Posts: 63
Default

Based on my limited understanding of how TopHat works, I would be very hesitant to use reads as short as 30bp. It seems to me you would have a pretty good chance of either the front or the back of the read (before or after the splice) mapping to random sections of the genome that way. For a small genome (like something microbial) that may not be too much of a problem. For anything comparable in size to human I would be extremely hesitant to trust any junction mapping without at least 40bp, and not very confident without 50+, as that would allow for a few mismatches before random matches became a serious problem.
mrawlins is offline   Reply With Quote
Old 10-18-2010, 05:53 AM   #3
Pejman
Member
 
Location: Switzerland

Join Date: Jul 2010
Posts: 23
Default

well, I'm working with human data, so ... but I'm not looking for new junctions, so I'm running it using predefined junctions from RefSeq and --no-new-junctions option, both for Tophat and Cufflinks. Now I've ran Bowtie -> cufflinks which is not recommended by the authors, and TopHat -. Cufflinks, which is recommended, but heavily fails on the alignment step. I'm gonna do some comparison, I'll keep you updated in case of any conclusive observations! Suggestions are welcome!
Pejman is offline   Reply With Quote
Old 10-29-2010, 03:29 AM   #4
waterboy
Member
 
Location: India

Join Date: Oct 2010
Posts: 14
Default

Hello Pejmen,

I am Shilp Purohit, and I am currently working with ABI SOLiD 3 plus sequencer. I am trying to install TopHat 1.1.2 (BETA) on my local machine. The "Getting started" manual suggests that I need to run following commands:

./configure
make
make install

However, the unzipped tar package doesn't have these files to execute, i.e. configure, make file and install file.

On the other hand, I unzipped tar package of TopHat version 1.1.0 and it worked absolutely fine and installed correctly just because the package consisted of these three files. But this version doesn't support the colorspace format of SOLiD.

Can you please suggest me as to how to install version 1.1.2? Any suggestions will be highly appreciated.

I am looking forward to having a reply soon.

Thanking you.
waterboy is offline   Reply With Quote
Old 10-29-2010, 04:12 AM   #5
Pejman
Member
 
Location: Switzerland

Join Date: Jul 2010
Posts: 23
Default

Hi
The last version I've used was 1.1.1 and that supports CS data. I recommend you to just use the precompiled versions, if you have problem with compiling. I checked the source files for 1.1.2, the configure file is there!
Pejman is offline   Reply With Quote
Old 10-29-2010, 07:45 AM   #6
JohnK
Senior Member
 
Location: Los Angeles, China.

Join Date: Feb 2010
Posts: 106
Default

How risky do you consider 50 bp reads?
JohnK is offline   Reply With Quote
Reply

Tags
solid, tophat

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 11:04 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO