Benchmarking bioinformatic tools for fast and accurate eDNA metabarcoding species identification

Area of Science:

• Environmental DNA (eDNA) metabarcoding
• Bioinformatic analysis
• Biodiversity assessment

Background:

• eDNA metabarcoding is vital for biodiversity assessment, but numerous bioinformatic tools exist with limited comparative performance data.
• Evaluating the speed and accuracy of these tools is essential for reliable species identification from eDNA data.

Purpose of the Study:

• To evaluate the performance of 13 bioinformatic programs and pipelines for analyzing fish eDNA metabarcoding data using the 12S mt rRNA gene.
• To compare program outputs based on sensitivity, F-measure, root-mean-square error (RMSE), and execution time.
• To develop and assess an optimized bioinformatic pipeline for fish eDNA analysis.

Main Methods:

• Utilized simulated mock communities and real fish eDNA metabarcoding datasets.
• Assessed 13 bioinformatic programs and pipelines focusing on the taxonomic assignment step.
• Compared program performance using sensitivity, F-measure, RMSE for read abundance, and execution time.
• Constructed and compared a novel pipeline against existing toolboxes (OBITools, Barque, QIIME 2).

Main Results:

• Significant differences were observed among programs, primarily in the taxonomic assignment step, affecting sensitivity, F-measure, and RMSE.
• Execution times varied considerably across programs and analysis steps.
• The developed pipeline and the Barque pipeline demonstrated superior performance across all tested indices.
• Real eDNA data analysis confirmed performance differences in taxonomic assignment and execution time.

Conclusions:

• The choice of bioinformatic program, particularly for taxonomic assignment, significantly impacts biodiversity estimates from eDNA metabarcoding.
• The developed pipeline and Barque are recommended as effective alternatives for analyzing fish eDNA metabarcoding data, especially with complete reference databases.
• Further research should focus on optimizing taxonomic assignment algorithms to improve the accuracy and reliability of eDNA biodiversity assessments.