PomBase: a Global Core Biodata Resource-growth, collaboration, and sustainability

Abstract

PomBase (https://www.pombase.org), the model organism database (MOD) for fission yeast, was recently awarded Global Core Biodata Resource (GCBR) status by the Global Biodata Coalition (GBC; https://globalbiodata.org/) after a rigorous selection process. In this MOD review, we present PomBase's continuing growth and improvement over the last 2 years. We describe these improvements in the context of the qualitative GCBR indicators related to scientific quality, comprehensivity, accelerating science, user stories, and collaborations with other biodata resources. This review also showcases the depth of existing connections both within the biocuration ecosystem and between PomBase and its user community.

Keywords: Schizosaccharomyces pombe; FAIR; Global BioData Coalition; Global Core Biodata Resource; MOD; biocuration; database; fission yeast; knowledgebase; model organism; model organism database; standardization.

Fig. 1.

Metrics to assess curation growth and coverage. a) Experimentally supported curated statements or annotations, sourced from the primary literature, show a consistent year-on-year growth (https://curation.pombase.org/pombe/stats/annotation). Major data types collected by PomBase are shown. b) Classifying proteins as functionally “known” or “unknown” is 1 way to describe the completeness of both curation and scientific knowledge (https://www.pombase.org/status/gene-characterisation). Here, we classify protein-coding genes as “known” or “unknown” based on GO biological process assignment to biologically meaningful grouping classes and show changes over time. We separate known into (1) “known from an experiment in S. pombe” (red / bottom band), where the process is completely or partially characterized in low-throughput experiments in S. pombe, and (2) “known from an inference from another species” (yellow / band second to bottom). We separate “unknown” into (1) “conserved unknown” (dark gray / band second from top), present in some species outside the Schizosaccharomyces clade, and (2) “species-specific unknown” (light gray / top band), restricted to the fission yeast clade.

Fig. 2.

New gene page summary and protein feature views. a) The gene page summary section has been redefined to allow users to set their preference to the existing JBrowse (genome browser) view or one of several new views. b) The new protein feature summary view uses the RCSB 1D Feature Viewer to display protein features (i.e. mutated residues, modifications, and structural features), illustrated here using the Cdc15 F-Bar membrane adaptor (https://pombase.org/gene/SPAC20G8.05c). c) The new protein feature “full view” also uses the RCSB 1D Feature Viewer and is accessible via a link from the summary view. This view provides a comprehensive display of engineered amino acid mutations (substitutions and deletions) linked to phenotype information. Mouse-over provides feature details. Illustrated here using Cdc15 protein F-Bar membrane adaptor (https://pombase.org/gene_protein_features/SPAC20G8.05c).

Fig. 3.

New structure views. Mol* (/molstar/), a web-based open-source toolkit developed by RCSB-PDB, is used to display protein structures from the PDB and AlphaFold. Illustrated here using the Lsm7 protein PDB entry as part of the Lsm2-8 complex (https://www.pombase.org/gene/SPCC285.12).

Fig. 4.

The Rhea reaction widget. The widget provided by the Rhea reaction database makes use of the GO to Rhea mappings to display reactions for all annotated GO activities.

Fig. 5.

Major knowledge dissemination pathways from PomBase. PomBase provides regular updates of curated data to major bioresources and repositories. We export manually curated gene structures to the ENA. ENA gene structures are used to provide the UniProtKB protein sequence (and subsequently used by InterPro resources and AlphaFold). We submit protein families and protein family updates to Pfam. As part of the GOC, we export functional annotation to the GO repository which is further disseminated to GOA and UniProtKB (and to NCBI and FungiDB, not shown). Phenotype data are exported to the Monarch Initiative knowledgebase. RNA features and GO curation are exported to RNACentral. The genome and nucleotide sequences and features are imported by Ensembl Genomes. Protein and genetic interaction data are submitted to BioGRID.

Fig. 6.

Current low-throughput curatable publication status by year published. The current curation status of the low-throughput publications identified as containing gene-specific experimental data (i.e. curatable) classified by publication date as of November 2023. Five categories are displayed: (1) admin curated (green): papers curated exclusively by PomBase staff who concentrate on legacy publications; (2) community curated (blue): curated papers with annotations contributed by authors (these are largely restricted to papers published since 2012 when Canto was launched); (3) uncurated, curator assigned: publications assigned to a PomBase curator but not yet curated; (4) uncurated, community assigned (red): assigned to a community member but not yet curated; and (5) uncurated, unassigned: not yet assigned for curation.