Re: [Inchworm-users] De Novo Assembly

[Brian J H. <bh...@br...>]

Hi Andrew,

Inchworm should work very well on squid, and strand-specific data
should be slightly better in this case than non-strand-specific data.
The strand-specific data helps mostly in the cases where you have high
gene density (some fungal genomes) and when transcripts slightly
overlap from opposite strands.

Since you don't have a genome, I suggest running the Inchworm assembly
output through CD-HIT with a 90% identity filter to pull out those
transcripts that are the highest quality assemblies.  This will
minimize the number of artifacts that fall through (assemblies based
on error-containing reads) and concentrate the results on the highest
quality transcripts.

In any case, if you have the option to go strand-specific, I highly
encourage you to do so.  Being able to properly differentiate between
antisense transcripts and sense-transcripts for genes is important and
can only be done properly with the strand-specific data.

Best,

-brian



On Wed, Nov 17, 2010 at 11:31 AM,  <and...@hu...> wrote:
> Hello-
>
> I'm a PhD student in Dr. Spencer Nyholm's lab at UConn. We're thinking of
> doing some strand specific sequencing and are interested in using inchworm
> for the assembly.
>
> Our animal model (a squid) has no sequenced genome, so I'm curious about
> how well inchworm assembles contigs without mapping the illumina reads to
> a genome. How do the resulting contigs compare with contigs assembled from
> unspecific strand reads? Can you give me some comparative statistics?
>
> Thank you.
> -Andrew
>
>
> ------------------------------------------------------------------------------
> Beautiful is writing same markup. Internet Explorer 9 supports
> standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
> Spend less time writing and  rewriting code and more time creating great
> experiences on the web. Be a part of the beta today
> http://p.sf.net/sfu/msIE9-sfdev2dev
> _______________________________________________
> Inchworm-users mailing list
> Inc...@li...
> https://lists.sourceforge.net/lists/listinfo/inchworm-users
>



-- 
--
Brian J. Haas
Manager, Bioinformatics Outreach, Genome Annotation and Analysis
The Broad Institute
http://broad.mit.edu/~bhaas