SEQanswers

Go Back   SEQanswers > Bioinformatics > Bioinformatics



Similar Threads
Thread Thread Starter Forum Replies Last Post
segment length in tophat ickou RNA Sequencing 4 05-21-2013 12:30 PM
BWA seed length parameter effects on speed and accuracy oiiio Bioinformatics 0 03-29-2011 09:05 PM
how to set seed length in Tophat sanwen Bioinformatics 2 08-10-2010 12:07 AM
Seed length in Maq sam253 Bioinformatics 0 04-27-2010 09:17 PM
TopHat limit length maria.b Bioinformatics 2 10-21-2009 11:50 PM

Reply
 
Thread Tools
Old 09-20-2009, 05:20 PM   #1
joseph
Member
 
Location: ca

Join Date: Feb 2008
Posts: 39
Default seed length in TopHat?

Is there a way to control the seed length in the TopHat command?
I know in Bowtie I can use -l/--seedlen or -3/--trim3.
Thanks
Joseph
joseph is offline   Reply With Quote
Old 09-23-2009, 02:44 PM   #2
sdriscoll
I like code
 
Location: San Diego, CA, USA

Join Date: Sep 2009
Posts: 438
Default

i believe you'd use the --segment-length <int> option. my data is 87bp and when left up to it's own devices Tophat splits those into 3 chunks of 25, 25 and 37. I tried a few times to set the --segment-length value to the length of my reads, 87, but it crashed Tophat each time (buffer overflow). I also set --segment-length to 40 which led Tophat to split my reads into 2 parts. that didn't crash it.
sdriscoll is offline   Reply With Quote
Old 09-23-2009, 03:22 PM   #3
joseph
Member
 
Location: ca

Join Date: Feb 2008
Posts: 39
Default

Actually, I heard from Cole that this parameter is controlled by TopHat, users cannot set it.
joseph is offline   Reply With Quote
Old 09-23-2009, 04:20 PM   #4
sdriscoll
I like code
 
Location: San Diego, CA, USA

Join Date: Sep 2009
Posts: 438
Default

users are not allowed to control the parameters sent to Bowtie because Tophat relies on a certain configuration and it has to generate a certain amount of stuff or else it's algorithms don't work. I thought about tweaking the Bowtie settings myself until I realised this from reading about Bowtie and then reading more about Tophat.

however, if you were to set --segment-length to the same length as your reads (in my case 87) then when Tophat passes the reads to Bowtie for alignment in the first stage it forces Bowtie to align the 87bp reads end-to-end. If you allow Tophat to split the reads up into smaller pieces it does so BEFORE Bowtie runs its initial alignments so it'll align reads of 25bp instead of the full 87 in my case. so in a way it is similar to the options you'd pass to Bowtie for seed length but in Tophat's case it applies to the alignment and the junction search.
sdriscoll is offline   Reply With Quote
Old 09-23-2009, 04:54 PM   #5
joseph
Member
 
Location: ca

Join Date: Feb 2008
Posts: 39
Default

What would be the difference between the two options (Tophat forcing Bowtie to do end-to-end mapping versus the default option of TopHat where the reads are split)? I am specifically interested in the effect on splice junction mapping.
joseph is offline   Reply With Quote
Old 09-23-2009, 09:23 PM   #6
sdriscoll
I like code
 
Location: San Diego, CA, USA

Join Date: Sep 2009
Posts: 438
Default

Without an answer directly from the author I'd say your best bet is to just run your data with those different options and compare the results. In fact I'm going to do this myself because I'd like to know more about how these settings alter the output as well.
sdriscoll is offline   Reply With Quote
Reply

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 10:26 PM.


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