Clustering/Binning Metagenomic contigs with MyCC and 16S sequences

[kopardev]

I have tried to use MyCC (http://www.nature.com/articles/srep24175) on samples with varied diversity and I always get very good accuracy between number of clusters and number of 16S sequences found in the metagenomic assembly. But, the problem is that all 16S sequences end up being part of 1 or 2 clusters. Why so? Can anyone explain this????

[Brian Bushnell]

16S is very highly conserved and is probably more similar to 16S from another species than it is to the rest of the genome, so it's natural that they would form a cluster. To do 16S clustering, you would need a 16S clustering tool, but you don't need one since you are working with assembled contigs rather than amplicons. I guess you probably want to figure out which 16S goes with which cluster... maybe you could accomplish that using paired read mapping information, or the cluster taxonomic assignments. I don't know of a tool that does it automatically.