Contradictory results comparing alpha diversity between two groups?

General Discussion
1

Hello, I am writing a paper related to the microbiome.

When comparing alpha diversity between two groups and observing that both Shannon and Simpson values are lower, how should it be interpreted?

As far as I know, a lower Shannon value indicates lower diversity, while a higher Simpson value suggests higher diversity.

Why would contradictory results be obtained in this case?

2

Hello @kanijuuuni,

As far as I know, a lower Shannon value indicates lower diversity, while a higher Simpson value suggests higher diversity.

Yes, higher values indicate greater diversity for both metrics. If I understand correctly, you have two samples one of which has a lower Shannon's and a lower Simpson's diversity metric than the other one. Here is the definition of Shannon's diversity used by qiime2, and here is the definition of Simpson's.

Why would contradictory results be obtained in this case?

I'm not sure I see the contradiction--is it possible you made a typo when describing the diversity metrics?

3 Likes
3

Thanks for your reply

I had a mistake describing metrics, Sorry for the typo!!

Lower Shannon value indicates lower diversity,
Higher Simpson value suggests lower diversity

I have two samples one of which has a lower Shannon's and a lower Simpson's diversity metric than the other one

5

Hello @kanijuuuni,

Lower Shannon value indicates lower diversity,
Higher Simpson value suggests lower diversity

According to the metrics that qiime2 uses, higher values actually indicates higher diversity for both. I believe the confusion is coming from a distinction between Simpsons D and Simpsons Index. Simpson's D (sometimes called dominance) does in fact indicate lower diversity with higher values. However, Simpson's index is defined as 1 - D (at least the definition used by qiime2), thus flipping the relationship. See here.

3 Likes