Dnabarcoder: An open-source software package for analysing and predicting DNA sequence similarity cutoffs for fungal...

Area of Science:

• Mycology
• Bioinformatics
• Molecular Ecology

Background:

• Fungal molecular identification and classification face accuracy challenges, especially in environmental metabarcoding.
• Current methods often use a single similarity cutoff, which is insufficient due to variable DNA marker variability.
• This limitation impacts the precision of fungal identification in large datasets.

Purpose of the Study:

• To introduce dnabarcoder, a novel tool for predicting local similarity cutoffs.
• To assess the resolving power of DNA barcodes for fungal sequence identification across different clades.
• To improve the accuracy and precision of fungal classification in metabarcoding studies.

Main Methods:

• Development and application of the dnabarcoder tool to predict local similarity cutoffs.
• Utilizing a fungal ITS DNA barcode dataset from the Westerdijk Fungal Biodiversity Institute.
• Classifying the UNITE database against the barcode dataset using both traditional and predicted cutoffs.

Main Results:

• Predicted similarity cutoffs varied significantly across fungal clades.
• Local cutoffs assigned fewer sequences but significantly improved accuracy and precision compared to traditional methods.
• Extracting the ITS region from ITS barcodes optimized taxonomic assignment accuracy.
• Full-length ITS, ITS1, and ITS2 showed similar species-level resolving power, with complete ITS superior at higher taxonomic levels.

Conclusions:

• dnabarcoder enhances fungal classification accuracy and precision in metabarcoding.
• Local similarity cutoffs are more effective than single, universal cutoffs.
• Optimizing DNA barcode region extraction and considering taxonomic level are crucial for accurate fungal identification.