Abstract

Motivation: The ongoing revolution in genome sequencing is delivering an unprecedented number of genome assemblies to global repositories, resulting in an overwhelming amount of data imported to UniProt in the form of proteomes. To manage this growth sustainably, there is a need for a systematic workflow to select the best proteomes. Results: We propose a novel pipeline for cellular organisms to select the best Reference Proteomes, i.e. those that best represent the protein space of a species. The pipeline uses a clustering algorithm based on MMseqs2 to select the minimum number of Reference Proteomes whilst maximising the representation of the protein space for each species. Additionally, we aligned our viral Reference Proteomes with the exemplar genome set defined by the International Committee on Taxonomy of Viruses. Because this method ensures that all species are represented with at least one Reference Proteome, the UniProt Knowledgebase increased the number of Reference Proteomes of 36% and covering 34% more species in the Tree of Life. The UniProt Knowledgebase will mainly retain proteins from Reference Proteomes and therefore this method reduces the overall number of proteins by 43%, leading to a more concise yet representative knowledgebase. ### Competing Interest Statement The authors have declared no competing interest. National Human Genome Research Institute, https://ror.org/00baak391 Office of the Director, https://ror.org/00fj8a872 National Institute of Allergy and Infectious Diseases National Institute on Aging, https://ror.org/049v75w11 National Institute of General Medical Sciences National Institute of Diabetes and Digestive and Kidney Diseases National Eye Institute National Cancer Institute National Heart, Lung, and Blood Institute National Institutes of Health, U24HG007822 European Molecular Biology Laboratory State Secretariat for Education, Research and Innovation