Dear All,
Let’s imagine we have made a nexus file of an alignment of 43 complete mitochondrial genome of Nymphalidae butterflies and reconstructed a Bayesian phylogenetic tree from it. Now for comparison, we want to reconstruct a Bayesian phylogenetic tree “only” from protein coding genes (PCG).
Is it possible to just find the approximate position of those 13 PCGs in the alignment and exclude everything else from the analysis using “exclude” command in MrBayes?
Or can you define those approximate positions as gene blocks and include them in your analysis using “include” command in MrBayes and leave the rest of the sequence alone?
What if you want to exclude 3rd codon position in the process? How do you deal with the Genes ND5, ND4, ND4L and ND1 that are in reverse orientation in the alignment?
I am asking this to make sure that there is no easy way of doing this and I have to go through 43 species and collect 13 genes from each mitogenome painstakingly and make 13 separate PCG alignments and concatenate them to make a single alignment file and do the Bayesian tree reconstruction based solely on PCG!
Thank you very much for your help and replies.
Roohollah Abbasi Hello! Interesting question. I had such a task, and I had to do it manually, because one species could have a shorter gene, another a longer one, and this, as you can understand, was important. And yes, some genes could even "jump" from the plus strand to the minus strand of the genome. It will be interesting to read the answers if anyone knows how to automate this process! Thanks for your question!)
If you have the complete mitochondrial genomes aligned, you can use the exclude command:
https://github.com/NBISweden/MrBayes/blob/develop/doc/manual/Manual_MrBayes_v3.2.pdf
And there are other ways to create a "mask" sequence in a multiple sequence alignment file, and then take columns from your alignment that are not masked.
https://academic.oup.com/mbe/article/30/3/689/1040880
I would definitely follow Brian's advice. He is a pro.
Still, there really is no easy way to do any large phylogeny. And the long tedious method you describe is at least useful in allowing you to know each of these sequences very well and to begin recognizing similarities and differences before you begin your analysis and choose your model of sequence evolution. These small features can have a very large effect on the accuracy of your final tree estimate. So you must be familiar with each sequence relative to the others. That is probably 95% of the work of phylogenetic analysis. Then plugging an informative alignment with a proper model is a cinch.
Having said that Protein and DNA sequences on Genbank are completely annotated with respect to the precise location along the sequence of the tRNAs, the control region and each of the protein coding genes, including which strand they are on. So this process must be possible and I am grateful that each sequence I download is the result of someone's annotations.
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology