Benjamin Dubois ^1* Mathieu Delitte ² Salomé Lengrand ² Claude Bragard ² Anne Legrève ² Frédéric Debode ¹

• ¹ Walloon Agricultural Research Centre, Gembloux, Belgium
• ² UCLouvain, Louvain-la-Neuve, Belgium

The study of sample taxonomic composition has evolved from direct observations and labor-intensive morphological studies to different DNA sequencing methodologies. Most of them these studies leverages the metabarcoding approach, which involves the amplification of a small taxonomically-informative portion of the genome and its subsequent high-throughput sequencing. RThe recent advances in sequencing technology brought by Oxford Nanopore Technologies have revolutionized the field, allowing enabling portability, affordable cost and long-read sequencing, thus therefore leading to a significant increase in the taxonomic resolution. However, Nanopore sequencing data exhibit a particular profile, with a higher error rate compared with Illumina sequencing, and existing bioinformatics pipelines for the analysis of such data are scarce and often insufficient, requiring specialized tools to accurately process long-read sequences.We present PRONAME (PROcessing NAnopore MEtabarcoding data), an open-source, userfriendly pipeline optimized for processing raw Nanopore sequencing data. PRONAME includes precompiled databases for complete 16S sequences (Silva138 and Greengenes2) and a newly developed and curated database dedicated to bacterial 16S-ITS-23S operon sequences. The user can also provide a custom database if desired, thus therefore allowing enabling the analysis of metabarcoding data for any domain of life. The pipeline significantly improves sequence accuracy, implementing innovative error-correction strategies and taking advantage of the new sequencing chemistry to produce high-quality duplex reads. Evaluations using a mock community have shown that PRONAME delivers consensus sequences showing demonstrating at least 99.5% accuracy with standard settings (and up to 99.7%), making it a robust tool for genomic analysis of complex multispecies communities.Conclusions: PRONAME meets the challenges of long-read Nanopore data processing, offering greater accuracy and versatility than existing pipelines. By integrating Nanopore-specific quality filtering, clustering and error correction, PRONAME produces high-precision consensus sequences. This brings the accuracy of Nanopore sequencing close to that of Illumina sequencing, while taking advantage of the benefits of long-read technologies.