This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Choosing K-mer size in Kmergenie for denovo assembly

How should we select the correct kmer size for kmergenie analysis? I have read in wiki page that 'largest k-mer size to consider'. In that case I am looking for valuable comments from you people.

next-gen assembly genome

Thank you 5heikki

But can you explain a little bit more. Are you saying to run with different k values less than max read length and take an average of them ?

I'm saying you should make assemblies with many different k-mer values and then in the end you merge all the assemblies into one. Also, you obviously can't have larger k-mer values than your read length, I mean, how would you e.g. get a 120-mer from a 90 bp read?

2 answers

There may be a misunderstanding here.

"Selecting the correct kmer size for kmergenie analysis" implies that kmergenie would perhaps require user assistance?

This is precisely the opposite, it is designed to choose k for you. Just run kmergenie without any parameter, and it will explicitly print what it thinks is a good kmer size. Most of the time, the value is accurate.

Additionally, for unusual cases, you can take a look at the generated HTML report to make a more informed decision.

Smaller than the length of your reads, larger than 20. Attempt many different values and merge the assemblies in the end.

Log in to answer this question.