Clustering Dna Sequences Using K-Means

Question

[Monzoor]

I want to cluster DNA sequences using oligo-nucleotide frequency vectors. Are there stand-alone implementations of k-means programs available for the same.

Answer 1

[Michael]

I gave an answer to a similar question here using R code.

If you replace hclust with kmeans then you already got there and R becomes a 'stand-alone solution' with a little script.

Try ?kmeans to see the available options. For very large datasets you can also try the Kmeans implementation in the amap package.

[Monzoor]

Thanks Michael. Will try the same and let you know.

Answer 2

[Prateek]

Not sure how different its is from k-means, but cd-hit is the one I use for clustering protein seqs - you can also use it for nucleotide seqs. It's an incremental clustering algorithm and is pretty fast.

site - http://www.bioinformatics.org/cd-hit/

user's guide - http://www.bioinformatics.org/cd-hit/cd-hit-user-guide.pdf

[Michael]

I think that is a different kind of 'clustering'