For sequencing data, one can do rarefaction analysis. My question was about different kinds of this analysis.
When doing a study that gives you nucleotide sequences of a certain functional gene, is it obvious to do this analysis on DNA level? Because what you are trying to figure out is what the minimum diversity in the sample is (based on for example Chao1) and how many additional sequences you still need to obtain the full diversity from the sample.
When working with a functional gene, you use the protein sequence. Due to degeneration of the code, multiple DNA sequences can give the same protein sequence. When making an OTU table based on proteins, rarefaction analysis can also be done. My question is, is this meaningful? In my opinion, all you get to know is how many of the functional diversity has been sampled, but you will never know how many additional sequences you need from your sample to obtain all this functional diversity right?
Thanks for the help!
Maybe with Species Diversity & Richness or EstimateS, you can do it.
Yes, I think is meaningful. By rarefying you homogenize sampling effort.
See attached. Hurlbert 1971, This is a good starting point for ecologists. Many papers since!
You might look at what is possible with the software GraphPad Prism 6, designed for the drug industry to evaluate dosages. I do not think this software includes a rarefaction routine, but may be useful to you otherwise. I had not considered how rarefaction analysis could be applied to evaluation of genetic diversity -but the concept of such an application seems sound. If the data are amenable to analysis via rarefaction, and an asymptotic curve can be produced, then you might be able to use differential calculus to define the point of greatest inflection on the curve, the point after which many additional samples add only limited additional diversity.
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology