FuzzyID2: A software package for large data set species identification via barcoding and metabarcoding using hidden...

Area of Science:

• Genomics
• Bioinformatics
• Ecology

Background:

• DNA barcoding and metabarcoding are crucial for biodiversity and ecological research.
• Large reference libraries present computational challenges for rapid species identification.
• Efficient data processing is needed for large-scale biodiversity assessments.

Purpose of the Study:

• To develop a faster and more accurate method for species identification from large DNA sequence datasets.
• To create a software pipeline (FuzzyID2) that enhances the efficiency of DNA-based species identification.
• To assess the reliability and accuracy of the new identification approach.

Main Methods:

• A two-step strategy combining Hidden Markov Models (HMM) for genus-level narrowing and minimum genetic distance for species identification.
• Development of the FuzzyID2 software pipeline using Python and C++.
• Incorporation of a fuzzy membership function to estimate the credibility of assignment results.

Main Results:

• FuzzyID2 achieved high mean accuracies: 98.60% for genus identification and 94.17% for species identification across diverse datasets.
• The method demonstrated significantly higher identification success rates compared to BLAST for simulated NGS data.
• Processing of datasets with tens of thousands of barcodes required only seconds per query assignment.

Conclusions:

• FuzzyID2 offers an efficient and accurate solution for species identification in biodiversity research dealing with large DNA sequence datasets.
• The developed approach addresses the time-consuming nature of global sequence library searches.
• This tool enhances the feasibility of large-scale biodiversity evaluations and ecological studies.