Assessing group differences in biodiversity by simultaneously testing a user-defined selection of diversity indices.

Area of Science:

• Ecology
• Bioinformatics
• Statistics

Background:

• Comparing diversity across ecological or metagenomic groups is common but challenging.
• Traditional diversity indices (e.g., Shannon, Simpson) have interpretation and comparison issues.
• Transforming raw indices into 'true' diversity measures can address these limitations.

Purpose of the Study:

• To introduce a novel statistical technique for comparing diversity between multiple groups.
• To enable simultaneous testing of user-selected 'true' diversity measures.
• To provide multiplicity-adjusted P-values to control false positive rates.

Main Methods:

• A resampling-based multiple testing procedure is employed.
• The method utilizes the Westfall and Young (1993) approach for P-value adjustment.
• The technique is implemented in the R package 'simboot'.

Main Results:

• The developed technique allows for robust comparison of diversity across multiple groups.
• It effectively corrects for multiple testing when evaluating various diversity indices.
• Ensures that the rate of false positives (type I error) is controlled.

Conclusions:

• This method offers a statistically sound approach to diversity comparison in biological studies.
• It overcomes limitations of traditional diversity indices, enhancing interpretability.
• The availability of the 'simboot' R package facilitates its application.