[BiO BB] 454/Newbler assembly finishing software?

[Dan Bolser]

2009/12/1 Edith Schlagenhauf <ediths at botinst.uzh.ch>:
> Hi,
>
> I've got a bunch of output files from a 454/Newbler de novo assembly
> for a 8 Mb bacterial genome, ie. various SFF/FASTA/ACE files.
>
> What are the most appropriate tools/software for viewing/evaluating the
> results and for finishing rapidly by automation (there are still several
> hundreds of gaps...)?
> There is also a cosmid end library available, and various related genomes.

For manual contig inspection and editing, the 'classic' choice is
still consed. I'm not sure but I think that it's about 130 years
old...

What you could try is mapping all your 454 reads against the related
genomes to get a second set of contigs. You can then combine your
de-novo and your 'mapping' contigs with a program like minimus2.

The cosmid library can be added by a scaffolder like bambus.

If you've got money to spend, you could do some low coverage large
insert size Illlumina sequencing and try to 'scaffold' the contigs
using that data.

Else you could try different assembly packages (i.e. Mira, Velvet,
CA-BOG, Phrap, etc.), and again try to combine your contig sets with
minimus2.

Also... (I've just come back from an assembly workshop ;-) You could
try walking in the Newbler parameter space and seeing if you can
'improve' the assembly (depending on how you want to judge that).

For assessing quality of the assembly, try viewing in Hawkeye, and
also running the AMOS validate pipeline.

Unfortunately, I don't know of any 'automatic finishing' software...
Consed will design primers for you though. Do you really need to
finish? Try annotating what you have.


Currently SeqAnswers seems the place to ask NGS related questions. But
don't forget to put bbb in your sig ;-)


Dan.

> TIA for your input,
> Edith
>
>
>
> ******************************************
> Dr Edith Schlagenhauf
> Bioinformatics
> Institute of Plant Biology
> University of Zurich
> SWITZERLAND
>
> e-mail: ediths AT botinst DOT uzh DOT ch
> ******************************************
>
>
>
> _______________________________________________
> BBB mailing list
> BBB at bioinformatics.org
> http://www.bioinformatics.org/mailman/listinfo/bbb
>