How to perform phylogeny analyses

Question

[mahimunquad92]

I want to do project on phylogenetics on hiv subtype-1. but how I start the project on this, what tools are apply on this?

[Michael]

Hi mahi, your post is very poorly written and researched, such posts are often closed quickly. Please rework your post and read some introductory materials on phylogenetics first.

[5heikki]

I would start by reading like 10 review articles on HIV and then follow from there to research papers, which are more close to your desired topic.

[prabhjotkaur3110]

What is actually diffrence between NJ tree and UPGMA tree.

Secondly, I want to construct NJ tree from genetic distance values. Please suggest me some online softwares, with their input data format, because I am facing many problems in data entry.

Answer 1

[Brice Sarver]

Phylogenetics is a complex field that requires a lot of background knowledge. You do not provide any information about your dataset so I cannot help, but I'll give you a super quick overview of how I would produce a publication-quality phylogenetic tree using likelihood-based approaches. I would recommend reading up on the literature.

1. Obtain a molecular dataset of genes of interest.
2. Perform a multiple sequence alignment (an inference of homology) using Mafft/Muscle/Clustal/a standard program.
3. Estimate a model of nucleotide sequence evolution.
4. Estimate a tree under that model using BEAST/MrBayes/PAUP*/Garli. If the dataset is too large and the run will not converge, try RAxML under GTR+G.
5. Estimate nodal support values by the nonparametric bootstrap/calculating posterior nodal support on a consensus tree.

Errors or improper/inappropriate settings at any of these steps will produce an unreliable tree. Push-button pipelines often produce suboptimal results.

I am happy to provide more detail if you (or others) need it.

[bioslayer]

Your answer is very well worded and give a round explanation of phylgenetics. I want to take a chance and ask about some tips. If I have a good quality alignment and want to estimate which sequences is more ancestral to the other one, which algorithm is more appropriate to answering that question ? A neighbor joining one or one that is based on UPGMA. I haven't really found satisfying answers in the literature. All I found was an explanation of the different tree building methods and nothing on the merits of choosing between methods to answer specific evolutionary questions.

Thanks again

[Brice Sarver]

Both NJ and UPGMA use distances to construct a tree. NJ trees are not necessarily rooted, whereas UPGMA (as it's a clustering approach) are. The biggest difference is that UPGMA assumes a constant rate of evolution across the lineages i.e., a molecular clock; because this is often violated in empirical datasets, this approach is usually considered sub-optimal.

Since the NJ tree will be unrooted, sequences/taxa will be polarized based on how you perform the rooting. The placement of the root is one of the most difficult parts of estimating a tree. Depending on your question, an unrooted NJ tree may be appropriate for what you need. Alternatively, I'd recommend estimating a high-quality tree using a Bayesian approach such as BEAST which places a prior on the age of the root.