I am new to this and I have just constructed a phylogenetic tree of 24 bacteria isolated from a kefir generated from the PCR-amplified bacterial 16S rDNA partial sequences (ca. 900 bp) based on the maximum likelihood method using Tamura-Nei model (Tamura and Nei, 1993) with 1,000 Bootstrap replications on MEGA 6.0 software after ClustalW alignment. The result is attached.
Q1: I got the score of '14' and '40' at the nodes, are these numbers too low to be convinced about the data? Do these numbers matter when it comes to publication?
Q2: Please see the picture. When the bacteria species of the same clade have the same branch length, does that mean they all have EXACT the same 16S rRNA sequences? Or they are just very closely related?
Q3: Please refer to the picture. Does it mean the organism with the branch length at the farthest right regarded as the newest species that have been evolved among all bacteria tested?
I'd really appreciate your answer/explanation to these questions to help me understand the phylogenetic tree.
Thank you in advance.
Dear Vijitra,
Keep in mind that the tree is unrooted unless you have prior information on the evolutionary history of the organisms, which you could use to place the root on the tree. Without that information, you cannot identify the direction in which time proceeds along branches in the tree, and thus cannot distinguish older from younger nodes.
You also need to make sure your sequence alignment is reasonable. The really long branch to the L. hircilactis sequence you ask about in Q3 makes me wonder if your alignment got corrupted and this sequence has been shifted out of alignment with the others. (This happens pretty easily, and I see this a lot with assignments I give to students!)
There are some excellent resources available for introducing the basics of phylogenetic analysis. I'd recommend Barry Hall's Phylogenetic Trees Made Easy as a starting point for understanding the answers to a lot of your questions.
14 and 40% bootstrap support is definitely too low to trust. Most people only report bootstraps >50%, and that is still considered to be very weak support. Most of your other nodes have >80% support - that's generally considered strong support. It's acceptable to leave the unsupported nodes in the graphic (rather than collapsing them), but reporting the numbers only emphasizes that those nodes have no support. Just be sure to report what you used as your cutoff point.
It's difficult to be certain from a graphic, but my guess is that yes, if the branch length=0 then those sequences are exactly identical. You should really remove identical sequences from the alignment before analyzing them since zero branch lengths can cause problems for some algorithms. On the other hand, since these are 16S sequences and there are likely to be large numbers of insertions/deletions in the alignment, it is possible that they are not actually identical, but appear so as an artifact of how you handled gaps in the analysis. If you used what MEGA calls a "complete deletion", that may result in large numbers of sites being ignored by the analysis, and the differences between those sequences may align with those gaps. You can often recover that information by using "partial-deletion" which only ignores sites which have a gap in a pairwise comparison.
Dear Vitrija;
Q1: Yes they are, normally this kind of bootstrap low values are ignored simply removing them from the tree. Normally in basal nodes of the tree the bootstrap values are lower than the rest. This could happen because your chosen marker its unable to resolve the deep phylogenetic relationships of your taxa, or maybe you can improve your analysis (for example as Laura commented about the gaps).
Q2: Exactly, those four sequences of L. lactis have the same sequence.
Q3: No, its position in the tree could means it share a common ancestor with all the other taxa represented in the tree (just like and outgroup). However, the low bootstrap values of the nodes which connect L. hircilactis with the rest of the tree can´t allow us to define the phylogenetic relationships between L. hircilactis and the rest of the taxa. As I said in Q1, probably you need to improve your tree (adding genetic markers, taxa, etc…)
Best wishes
Dear, Vijitra
Q1 the bootstrap value 14 and 40 are weak. statistically significant bootstrap is a bootstrap from 95% (alpha = 0.05)
Q2 they are just very closely related, but it's possible the two clad have the same sequence.The sequence you can check by comparing the sequences by appropriate software
Q3 This means that this is the most different of the bacteria group study. it's a group out
Dear Vijitra,
Keep in mind that the tree is unrooted unless you have prior information on the evolutionary history of the organisms, which you could use to place the root on the tree. Without that information, you cannot identify the direction in which time proceeds along branches in the tree, and thus cannot distinguish older from younger nodes.
You also need to make sure your sequence alignment is reasonable. The really long branch to the L. hircilactis sequence you ask about in Q3 makes me wonder if your alignment got corrupted and this sequence has been shifted out of alignment with the others. (This happens pretty easily, and I see this a lot with assignments I give to students!)
There are some excellent resources available for introducing the basics of phylogenetic analysis. I'd recommend Barry Hall's Phylogenetic Trees Made Easy as a starting point for understanding the answers to a lot of your questions.
Thank you so much for all suggestions. I think I understand it much better now.
- Badges
- Science topic
- Similar topics
- Biological Science
- Evolutionary Biology
- Phylogenetics