gmmDenoise: A New Method and R Package for High-Confidence Sequence Variant Filtering in Environmental DNA Amplicon...

Area of Science:

• Ecology and evolutionary biology
• Genetics and genomics
• Bioinformatics and computational biology

Background:

• Monitoring genetic diversity is crucial for understanding population ecology and evolution.
• Traditional tissue sampling for genetic analysis is often difficult in wild species.
• Environmental DNA (eDNA) metabarcoding offers a promising alternative but is hampered by spurious sequences.

Purpose of the Study:

• To develop a novel amplicon filtering approach to eliminate spurious amplicon sequence variants (ASVs) in eDNA metabarcoding data.
• To improve the reliability of genetic diversity assessments and population genetic inferences from eDNA data.

Main Methods:

• Simulated eDNA metabarcoding processes to analyze read count distributions of true ASVs and PCR-generated artefacts.
• Developed a Gaussian mixture model-based approach to estimate abundance distributions and determine a statistical threshold between true and false-positive ASVs.
• Implemented the approach as an R package named gmmDenoise.

Main Results:

• The gmmDenoise approach effectively eliminates spurious ASVs from eDNA metabarcoding data.
• Evaluated using single-species datasets with known true ASVs, demonstrating reliable identification.
• Applied to community metabarcoding datasets, enabling robust intraspecific diversity estimates and population genetic inferences.

Conclusions:

• gmmDenoise significantly enhances the accuracy of eDNA metabarcoding for population genetic studies.
• The R package provides a valuable tool for researchers studying genetic diversity in natural populations.
• This method overcomes key challenges in eDNA analysis, paving the way for more reliable biodiversity monitoring.