Re: [wgs-assembler-users] pacBioToCA spec parameter optimization

[Walenz, B. <bw...@jc...>]

I think most of the time will be spent in overlaps, and this page should
help:

http://sourceforge.net/apps/mediawiki/wgs-assembler/index.php?title=RunCA#OV
L_Overlapper

Or just confuse.  I've been told that the page is a bit obtuse for
non-hackers, but it's all we have right now.

In general 4 cores and 8gb memory works great, for 'normal' genomes.  There
are a few more options that you can fiddle with to optimize memory loading,
but the defaults -- once you work through the page above -- should work
reasonably well (I hope).

Happy to help more if you want to get more specific.

b




On 8/14/12 9:26 AM, "Thomas Hackl" <tho...@un...> wrote:

> Hi,
> 
> I would like to use the pacBioToCA correction pipeline on data for a
> bigger genomes (Gb size), hence computation time quite concerns me.
> Judging from the *.spec files, there are quite some parameters to tweak.
> We have machines, 40/80 cores, 200/500Gb memory, and I was wondering if
> you could give me some advice on how to modify the spec file parameters
> to make optimal use of this potential.
> 
> Thanks
> Thomas
>