How much RAM is recommended for "mothur" to run cluster for grouping the sequences in OTU ?
The required memory will scale with the number of pairwise distances, as Mothur loads the entire distance matrix into memory.
How much data do you have, and what hardware do you have access to?
The distance matrix generated contains 18 Go and I use my computer windows 7 (64bit) with4GB memory
The distance matrix contains 18 Gb? Well then you definitely need more than 4 Gb. A process can't access all of the memory available on the machine due to usage by the OS and other necessary processes, so I would say you should probably find a machine with 32 Gb of memory. You don't quite need that much but usually you don't see anything in between 16 Gb and 32.
thank so very much for your answers and I'll consider your advice in seeking machine with 32 Gb of memory
About this question, I have a solution that I would like to share with you. For those who use the Mothur software and if you have obtained after the formation of the distance matrix a large file and you could not turn the cluster command to grouping the sequences in OTU . In this case, you can choose the threshold present (cuttof) in the command below according on the capacity of your computer: dist.seqs (fasta = datafinal.fasta, cutoff = 0.10, processors = 4). In my situation I used cutttof = 0.10 instead of cuttof = 0.15 and it worked well
I wish you good luck
- Badges
- Science topic
- Similar topics
- Biological Science
- Biota
- Eukaryota
- Fungi