search

adw96 / DivNet

diversity estimation under ecological networks
83 stars 18 forks source link

feature request: allow random effects in testDiversity #30

Closed yanxianl closed 5 years ago

yanxianl commented 5 years ago

Hi,

I just read the recent works from your group on estimating microbial diversity and relative abundance with measurement errors and decided to use the methods proposed. I was not aware of the pitfalls on using the observed diversity metrics as if they were the real ones until I read the papers. Great works that you have done!

I want to compare the alpha diversity among groups with random effects. While I can use the breakaway::betta_random to compare the species richness , similar functions for comparing the Shannon and Simpson index are missing in the Divnet package. It'd be great to also allow modelling random effects in the testDiversity as well.

Meanwhile, can I use betta_random to do hypothesis testing for Shannon and Simpon index?

Yanxian

adw96 commented 5 years ago

Dear Yanxian --

I apologise that I did not see your question earlier. Indeed I recommend breakaway::betta_random for random effects in the absence of this functionality in DivNet at this time.

Cheers,

Amy

adw96 commented 5 years ago

Hi Yanxian -- I just made a new vignette in breakaway that talks through fixed and random effects models for diversity. It includes both species richness (estimated using breakaway) and Shannon and Simpson diversity (estimated using DivNet).

Check out the vignette here:

https://github.com/adw96/breakaway/blob/master/vignettes/diversity-hypothesis-testing.Rmd

😸

yanxianl commented 5 years ago

Thanks! I'll read the new tutorial.