Aim:
Are the features that we are interested in more conserved in a set of virus strains than expected?
Input:
1) 100 virus genomes (strains of same virus) as pairwise aligned sequences (no MSA).
2) Features we are interested in as locations in one of the genomes (e.g.: GFF).
3) As 2) but randomly selected regions of same size as 2).
Intuitively, I would just calculate the differences for these features (2) over all genomes and compare the %nucleotide changes to random selected parts (3).
How should this actually be done according to the current state of the art and are there any tools to do this calculation?
See:
Genome Res. 2001 Jan; 11(1): 43–54, Genome-wide Function Conservation and Phylogeny in the Herpesviridae
I hope this helps
Thank you Andrew, this looks very suitable.
Perhaps the approach is a bit old and might have matured over the last decade. Anything more recent?
I would calculate the Shannon Entropy from a MSA. Also, covariant positions can be calculated from Entropy values using the Mutual Information formula. See: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4676106/
I think what you want is to calculate some measure of phylogenetic "signal", is that right? If so, how many strains do you have? Do you have some kind of phylogeny?
Probably a signal is what I am looking for. There are about 100 strains all from Zika virus and I am not sure whether there is a phylogeny but given the hype, there should be.
Yes, apparently there's one.
https://wwwnc.cdc.gov/eid/article/22/5/16-0065-f1
- Badges
- Science method
- Similar topics
- Biological Science
- Evolutionary Biology
- Phylogenetics
- Phylogenetic Analysis