On May 5, 2013, at 23:59 , raw937@xxxxxxxxx wrote: > I have a few metagenomes were I have data with all three platforms. > I have filtered the reads for low quality, phiX, dimers etc. Ummm … it is very rare that outside filtering produces a better assembly than filtering with MIRA. Well, to be honest: phiX is something which will come only in 3.9.16, but apart that: Illumina data should not be filtered or quality clipped by other tools, MIRA does a much better job. For 454 data one should take the clippings from the Roche software and then let MIRA do the rest. > What are the command parameters for such an assembly? I have them still in > fastq format. You should take 3.9.15 (or better, wait for 3.9.16 out RSN). Then either assemble in "genome" or "est" mode, depending whether your samples are real metagenome or simply clean cultures of some different strains. You'll find examples for setting up manifest files in the manual. > I have 20million HiSeq100PE, 2million MiSeq250PE and a million 454Ti for > about 8 samples. How much memory will I need? Unknown, highly dependent on your data. As a rough approximation I'd say: not below 48 GiB, but I may be wrong there. > And, how many trends should I give mira? As I want to assemble it rather > quickly. Rather quickly is always subjective, 3.9.16 will again be a bit faster than 3.9.15 for that many reads. Take something between 8 and 64 threads (if you have that many cores). 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