Re: [Useq-users] How to speedup the process of building junction library

[David N. <dav...@gm...>]

Hello Michael,

I break up the transcript file by chromosome and run each separately on a
cluster.   Another speed up is to set the max number of junctions allowed to
10000 instead of the default 100000.

Did you see the RNASeq extended junction guide?
http://useq.sourceforge.net/usageRNASeq.html

We've built many transcriptomes so take a look here to see if there is
something you can use.
https://bioserver.hci.utah.edu:443/gnomex/gnomexFlex.jsp?topicNumber=4

-cheers, D

-- 
David Austin Nix, PhD
Huntsman Cancer Institute
Dept. OncSci University of Utah
Bioinformatics Shared Resource
HCI 3165 (801) 587-4611
dav...@hc...


From:  michael xu <mic...@gm...>
Date:  Fri, 19 Oct 2012 20:35:43 -0500
To:  USeq <use...@li...>
Subject:  [Useq-users] How to speedup the process of building junction
library

Hi David,

Thank you for developing a very useful tool USeq!

I am using the USeq to build a novoalign index file with junction library,
but I found it takes a very long time (> 5+ hours or longer) to build a
junction library.
Until now, I still have not finish one junction library for human 19 on a
very powerful cluster.
Do you have any suggestion on how to build a junction library quickly?

Besides, I know I can download a library from your website with 50nt, but
the length of read in my dataset is 51nt. I am not sure if I can use the
same junction library build by you even the length of read is 51nt. Suppose
that -r should set "length of read - 4".

Your response would be greatly appreciated!

Michael
----------------------------------------------------------------------------
-- Everyone hates slow websites. So do we. Make your web apps faster with
AppDynamics Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_sfd2d_oct________________________________________
_______ Useq-users mailing list Use...@li...
https://lists.sourceforge.net/lists/listinfo/useq-users