cloud computing sever for DNA sequence assembly

[seb567]

For more memory for cloud computing server, you can ask and learn more about here erp on cloud

You may want to give a try to the SSD instances (hi1.4xlarge) and with that maybe a swap file on these will help for software requiring tons of memory.

Otherwise, I suggest you use Ray on cc2.8xlarge instances -- these have 16 cores, 32 threads, 60 GiB ram and 10 Gigabit Ethernet.


-Sébastien

[ramcob]

For more memory for cloud computing server, you can ask and learn more about here erp on cloud

[hengnck]

Thanks, Pseudonym. I'll install and try Gossamer out when the MIRA assembly finishes... it's only been 6 *days* continuously chugging away - two passes and another 2.36 million reads to go. I'll give you feedback on how Gossamer deals with Ion Torrent data.

[Pseudonym]

@Pseudonym - does Gossamer work on a 2008-vintage Intel Core 2 Quad 8200 with 8 GB of RAM (running Bio-Linux 6) to de novo assemble Ion Torrent Ion-318 shotgun data (480 Mbp)? The data is from an oral bacterium, probably 2.0-2.2 Mbp in size. Your paper is most interesting.

Thanks!

Memory won't be a problem at all; 8 GB is more than enough to assemble anything smaller than 250Mbp or so. The only catch I can see is that as far as we know, nobody has ever tried Gossamer on Ion Torrent data, and we don't know what its error characteristics are, so it may need a little fine-tuning to get good results.

Feel free to PM me if you need any help.

[hengnck]

@Pseudonym - does Gossamer work on a 2008-vintage Intel Core 2 Quad 8200 with 8 GB of RAM (running Bio-Linux 6) to de novo assemble Ion Torrent Ion-318 shotgun data (480 Mbp)? The data is from an oral bacterium, probably 2.0-2.2 Mbp in size. Your paper is most interesting.

[Pseudonym]

Again, plugging one's own project, Gossamer may be able to handle that volume of data on your existing hardware. Depending on your existing hardware, of course.

[krobison]

Another option is to use Ray, a de novo assembler that can spread the job across multiple nodes on a cluster.

For setting up such a cluster on Amazon, STAR::Cluster can be quite useful.

[kadircaner]

You might want to look here, if you are still searching:

http://ged.msu.edu/angus/metag-assembly-2011/running-trinity.html

[ians]

I've been looking for more memory as well.
As far as pre installed software, you could always boot an image of biolinux.

[jjjscuedu]

Dear all,

Do you know some famous cloud computing sever for DNA sequence assembly?

Because my sequence dataset is too large: 450 million for one end, 900 million for pair end.

I have checked the cloud computing on Amazon. However, for high memory instances, they just provide 68.4GB, it is not enough.

In addition, I want these sever have already installed some assembly software like Trinity, AbySS, velvet.

Do you know such sever?

Thanks!

Jingjing