optimal K mer length in MetaVelvet and command line to calculate N50

[shafiqaazmi]

Hai

I would like to assemble my metagenomes using MetaVelvet. I am planning to use VelvetOptimiser to determine the optimal K mer length first, then the determined K mer length will be used in my MetaVelvet assembly.

From my reading, the maximum K mer length for the VelvetOptimiser is 100, whereas the maximum K mer length can be used in MetaVelvet is 63.

Thus, I have to try several K mer length lesser than 63 in VelvetOptimiser right ? So that optimal K mer length could be used in MetaVelvet. Please correct me if I was wrong.

Secondly, may I assume that the number of nodes stated in the output file of VelvetOptimser is equal to number of contigs ? And when I proceed the assembly using MetaVelvet, I may assume that the number of nodes in the metavelvetg.LastGraph-stats.txt is equal to number of scaffolds of my metagenomes ?. What I had understood was MetaVelvet could reconstruct relatively low-overage genome sequences as scaffolds.

Lastly, do you mind to share any command line to calculate the length of N50 from my metavelvetg.LastGraph-stats.txt in linux perhaps or R statistical package ?. Looking forward for your response.

Thank You

[mastal]

Hai

From my reading, the maximum K mer length for the VelvetOptimiser is 100

Kmer lengths should only be odd numbers.

The maximum kmer length depends on the MAXKMERLENGTH that velvet was compiled with, and obviously on the length of your reads.

You can find out what the MAXKMERLENGTH is by typing 'velveth' without any parameters. You can recompile velvet if you want to use a higher kmer length.

Secondly, may I assume that the number of nodes stated in the output file of VelvetOptimser is equal to number of contigs ?

Yes.