Two-stage microbial community experimental design.

Area of Science:

• Microbiology
• Bioinformatics
• Ecology

Background:

• High-throughput sequencing of microbial communities, like 16S ribosomal RNA gene sequencing, is common.
• Two-stage study designs, involving initial surveys and follow-up analyses, are used in ecology but underexplored in microbial investigations.
• Ad hoc sample selection can introduce bias in microbial community studies.

Purpose of the Study:

• To develop and validate purposive sample selection methods for two-stage microbial community studies.
• To evaluate the accuracy and impact of different selection criteria on follow-up sample characteristics.
• To provide a software tool for implementing these selection strategies.

Main Methods:

• Developed several purposive sample selection methods based on biological criteria for two-stage microbial studies.
• Evaluated selection techniques using simulated data and experimental data from the Human Microbiome Project (318 paired 16S amplicon and metagenomic samples).
• Assessed criteria including selecting typical, clade-specific, rare, diverse, extreme, or discriminating communities.

Main Results:

• All developed criteria successfully identified samples with properties retained in follow-up analyses.
• Some selection criteria, like diversity maximization, resulted in follow-up samples that were not representative of the original survey population.
• Selecting intentionally representative samples minimized differences between the initial survey and the follow-up sample set.

Conclusions:

• Purposive sampling methods can be effectively applied to microbial community studies for optimized two-stage designs.
• Careful selection is crucial to ensure follow-up samples accurately represent the broader microbial community surveyed.
• The microPITA software facilitates the implementation of these validated sampling strategies for microbial research.