I have OTUs sequence generated from 16s V4 clean reads, and want to build maximum likelhood tree with selcted taxonomy 16s refseq from Sliva database.


However, the 16s sequences are always around 1500nt while my OTUS are around 200nt, which is not good for multiple sequence alignments. My question is how to select V4 from the 16s refseq.

  • $\begingroup$ Welcome to the site. The V4 and so are known regions of the 16S gene, you'll need to know where are they and then extract them to align with your OTU. But probably this is already done when deciding if it is an OTU or not. $\endgroup$ – llrs Jul 31 '19 at 10:01
  • $\begingroup$ Any chance of explaining why Sliva and not NCBI please? $\endgroup$ – M__ Jul 31 '19 at 12:43
  • $\begingroup$ Ok I'll explain, Sliva is a database tied to commercial (?) software. All the data on Sliva will be on NCBI. I can explain how you do this complete calculation from first principles, for your 200 BP using NCBI. The alignment (which is what Sliva gives) isn't the computational demanding algorithm, it's the tree. If you are wired to Sliva how ever ... That's more work (which someone else might do) $\endgroup$ – M__ Jul 31 '19 at 19:32
  • $\begingroup$ Oh wait the alternative is a blockwise deletion ML phylogeney, depends on the indel structure of the other taxa but that would work $\endgroup$ – M__ Jul 31 '19 at 19:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.