[mira_talk] Re: Regarding Time-estimate of mira
- From: bio5yz <bio5yz@xxxxxxxxxxxxxx>
- To: mira_talk@xxxxxxxxxxxxx
- Date: Tue, 21 Dec 2010 13:19:59 -0700
Not sure if this is helpful, but I made a recent 3.5 million 454 assembly
(~350 bp ave length) that took 10 days. This was a transcriptome assembly at
accurate setting.
Our system also has 512 GB memory, 4 hexacore processors (I believe).
Number of processors doesn't seem to affect the assembly too much, as I
believe it is only parrallelized during the skimming process. NFS is a bad
idea; my first run stalled and had to be killed. Is your run log file 500
GB or the log directory in the assembly folder?
Michael
On Tue, Dec 21, 2010 at 3:09 AM, Ganga Jeena <ganga.jeena@xxxxxxxxx> wrote:
> I have to do this assembly for 8 million reads454 in master-node/home
> within 230 GB space. How should I work it?
> The log file generated is already 500 GB.
> Is there some way | option to avoid log file generation and other
> non-result files ?
> Regards,
> Ganga Jeena
>
>
>
> On Tue, Dec 21, 2010 at 10:22 AM, Ganga Jeena <ganga.jeena@xxxxxxxxx>wrote:
>
>
>>
>> ---------- Forwarded message ----------
>> From: Bastien Chevreux <bach@xxxxxxxxxxxx>
>> Date: Mon, Dec 20, 2010 at 11:54 PM
>> Subject: [mira_talk] Re: Regarding Time-estimate of mira
>> To: mira_talk@xxxxxxxxxxxxx
>>
>>
>> On Montag 20 Dezember 2010 Ganga Jeena wrote:
>> > I am trying to do assembly of 8 million reads with mira
>> > Configuration of system is as follows:
>> > 512 GB RAM
>> > 8 hexa-core processors
>> >
>> > miramem gives me a memory estimate of 66.6 GB.
>> > How to get the time-estimate for assembly of data on my system.
>>
>>
>> It is impossible to estimate the time needed, there are too many
>> sequencing
>> project depending variables which make things absolutely unpredictable.
>>
>> Sorry, you'll have to try it out with your data.
>>
>> B.
>>
>> --
>> You have received this mail because you are subscribed to the mira_talk
>> mailing list. For information on how to subscribe or unsubscribe, please
>> visit http://www.chevreux.org/mira_mailinglists.html
>>
>>
>
Other related posts:
