Oxford Nanopore Technologies (ONT) workflows attach synthetic adapters and barcode oligonucleotides to DNA fragments for priming and multiplexing. Although these sequences are normally removed during basecalling or post-processing, incomplete trimming can embed non-biological fragments into genome assemblies deposited in public repositories. A systematic sequence-similarity screening of the GenBank nucleotide search database identified 5,782 unique accessions containing 13,734 Oxford Nanopore Technologies (ONT) adapter and barcode fragments embedded in bacterial, viral, and eukaryotic assemblies. Most contaminating sequences correspond to rapid-ligation and ligation-based adapters, reflecting incomplete trimming in ONT workflows. Multiple ONT synthetic sequences were often detected within a single accession, consistent with protocols employing several adapters and barcodes. Compared with GC-matched randomized control sequences, the original ONT sequence catalogue yielded significantly more GenBank matches, supporting that the observed signal exceeds chance similarity. This ONT-specific contamination recurs across GenBank records and exhibits structured patterns across taxonomic groups and project types. Implementing ONT-aware screening during public database submissions and curation is critical to prevent the ongoing propagation of synthetic sequences and safeguard genomic data integrity.
Martino et al. (Mon,) studied this question.