Semi-automated sequence curation for reliable reference datasets in ITS2 vascular plant DNA (meta-)barcoding.

Area of Science:

• Botany
• Genetics
• Bioinformatics

Background:

• Accurate DNA reference sequence datasets are essential for reliable taxonomic identification in DNA (meta)-barcoding, particularly for the Viridiplantae kingdom.
• Existing reference datasets often contain taxonomic and other errors due to reliance on general public databases like GenBank.

Purpose of the Study:

• To construct curated DNA reference datasets for the ITS2 barcoding marker of vascular plants.
• To develop a robust and updatable pipeline for automated and manual curation of these datasets.
• To enhance the accuracy and reliability of plant DNA reference databases for taxonomic identification.

Main Methods:

• Development of a pipeline script incorporating an automated curation stage with five filters.
• Manual taxonomic correction of misclassified taxa within the datasets.
• Manual addition of newly sequenced species to enrich the reference databases.
• Application of the pipeline to create a global dataset, a European crop dataset, and 27 EU country-specific datasets.

Main Results:

• The curation process identified and discarded 13% of imported sequences (7% of species) due to errors.
• 259 new sequences were manually added to the global dataset.
• The curated global dataset now contains 307,977 sequences representing 111,382 plant species.
• The pipeline facilitates efficient updating of curated datasets.

Conclusions:

• The developed pipeline significantly improves the quality and accuracy of vascular plant ITS2 reference datasets.
• The curated datasets provide a reliable resource for DNA barcoding and taxonomic identification in plants.
• This work addresses a critical need for high-quality reference data in plant science and conservation efforts.