When editing mitochondrial COI barcode sequences for submission, the standard practice is to follow the Barcode of Life Data System (BOLD) guidelines, which specify the use of “N” to denote any ambiguous base rather than retaining the full set of IUPAC ambiguity codes (e.g., R, Y, W, S). This ensures compatibility across databases and analytical pipelines, since many downstream tools in DNA barcoding and species delimitation expect either unambiguous bases (A, T, C, G) or “N” for uncertain positions. Using other IUPAC symbols may cause alignment or parsing errors and reduce interoperability when data are shared across repositories.

From a bioinformatics standpoint, limiting ambiguous calls to “N” helps standardize sequence representation and avoids interpretational inconsistencies, particularly when generating consensus sequences from bidirectional chromatograms. While IUPAC codes technically carry more information (e.g., distinguishing between A/G or C/T uncertainty), this added detail rarely translates into meaningful taxonomic resolution in barcoding workflows. Since the COI barcode relies on standardized sequence length and format, even a few ambiguous bases coded as “N” generally do not affect identification unless ambiguity occurs in large stretches.

Therefore, the recommended approach is to replace all ambiguous base calls with “N” during sequence editing and submission to BOLD, rather than using other IUPAC codes. If you want to preserve the additional ambiguity information for research purposes (e.g., for phylogenetic analysis or population genetics outside of BOLD), you can maintain a separate internal dataset with full IUPAC codes. This way, you comply with BOLD’s submission standards while retaining the option to analyze more nuanced sequence variation in your own workflows.

Kamyar Amirhosseini In mitochondrial COI barcoding workflows, particularly when preparing sequences for submission to BOLD (Barcode of Life Data Systems), it is recommended to use only the ambiguity code 'N' for uncertain bases, rather than the full range of IUPAC ambiguity codes (such as R, Y, S, W, K, or M). This is because the mitochondrial COI gene is haploid and uniparentally inherited, so it should not exhibit heterozygosity. The presence of multiple ambiguous base calls typically indicates sequencing errors, contamination, or the presence of nuclear mitochondrial pseudogenes (NUMTs), rather than true biological variation. As such, the use of extended IUPAC codes can interfere with downstream applications like sequence alignment, distance-based species identification, and phylogenetic analysis, and they are generally not accepted by BOLD.

When editing consensus sequences, the best practice is to carefully review the chromatogram or basecall data. If the correct base cannot be confidently determined, it is appropriate to assign an 'N' at that position. However, if a base is clearly supported by the majority of high-quality reads, it should be assigned accordingly. In cases where there is consistent and unresolvable ambiguity, it is advisable to resequence the region or exclude the affected portion from the final barcode. Ultimately, keeping the sequence as clean and unambiguous as possible ensures compliance with BOLD guidelines and improves the reliability of species identification.

H.A Adewuyi Anuraj Nayarisseri Thank you both very much for your replies. I also agree that annotating base ambiguities using the code N makes more sense, particularly when working with protein-coding mitochondrial genes like COI.

So thanks again for your contributions. I now have more confidence in my approach!

Just one more thing: do any of you have any publications, handbooks, guidelines, etc. at hand (basically any source that can be cited!), that clearly state/recommend N base calling for handling ambiguity in barcode sequences? If yes, I'd greatly appreciate it if you would share them with me.

Cheers!