I did nearly full 16S sequencing (1382 bps) of my isolate. According to blast result in EZ taxon data base the isolate showed 98.5% similarity to Kocuria polaris but when I drew phylogenetic tree based on Maximum likelihood and Neighbor joining algorithms, it constituted a monoclade with Kocuria aegyptica. I was wondering if any one could give me any explanation on this?
You may check the results of blast. Blast is "Basic Local Alignment Search Tool" which means it may only report sub-regions of highly similar sequences. However,a phylogenetic tree should be constructed based on global alignment algorithm considering the full sequences (such as clustal).
Did you trim the hanging ends of the alignment? It can give problems.
How's the topology of the rest of the tree? Does it match with the expected? For example the one in the paper linked below.
Article Kocuria atrinae sp. nov., isolated from traditional Korean f...
Dear Xabier Vázquez Campos,
Thank you for your reply. The attached you can find the situation of my isolate(contig1) among other kocuria type strains.
Well, the clade including your isolate and K. turfanensis is not that well supported (51). The tree is consistent with published material, except for the deepest blanches. And K. polaris is 99.5% identical to K. rosea, way closer than to your strain
What was the similarity score with for example, K. turfanensis?
What did you use to align the sequences?
The pairwise similarity between my isolate and Kokuria turfanensis in EZ taxon is 98.26% . I used MUSCLE for aligning the sequences implemented MEGA 6 version.
The top-scoring BLAST hit is not automatically the closest neighbor. (http://link.springer.com/article/10.1007/s002390010184) Whenever possible, generate and work off of a good quality phylogeny.
I suggest a manual aligning, using your chromatograms and your sequences, then run the ML & NJ trees, maybe you'll have some different trees and/or more well supported clades.
I would say that if you are within the Kocuria clade, that is about as good as you can get with just a 16S rRNA gene sequence. Defining species within that clade, based only on 16S, looks a little sketchy in my opinion. See the attached alignment and tree. Add your sequence to this alignment, and see what your tree looks like.
I adhere to Brian's answer. In adition, BLAST reports the Maximal-scoring Segment Pairs, whereas ML programs use the whole alignment, which could also explain this small disagreement.
Dear Brian,
I followed your suggestion. here is the results. Would you please tell me your opinion ???
Is that your "contig1" right with the DQ520814 Micrococcaceae bacterium NR120
sequence? If so, I don't think it would be considered to be a typical Kocuria.
My tree has JX293352_Actinomycetales_bacterium_enrichment_culture
outside the Kocuria clade, whereas your tree has that sequence sort of splitting the Kocuria into two clades. I am not sure why that would happen just by adding your contig1 to the alignment. Did you re-align?
Here, I added my sequence to pre-aligned sequences using MUSCLE and drew the tree using NJ method
- Badges
- Science method
- Similar topics
- Biological Science
- Evolutionary Biology
- Phylogenetics
- Phylogenetic Analysis