Fig. 4: Functional significance of Cell Cycle-Dependent (CCD) proteins.

A Fisher’s exact test Gene Ontology (GO) analysis of CCD proteins. The top three significantly enriched terms for each category (Reactome, Biological Processes, Molecular Function and Cellular Components) are shown. B The oscillation of CCD proteins within the same cellular localisation or complex. Scatter plot showing the curve fold change median (global oscillation of proteins within each complex) and standard deviation (global variance in oscillation within each protein group) for protein sets grouped based on Gene Ontology-defined protein localisation (left panel) or proteins within the same complex obtained from CORUM and Complex Portal complexes^21,22 (right panel). C Schematic of CCD proteins localised at the replication fork, Anaphase Promoting Complex/Cyclosome and kinetochore. Protein colour indicates protein abundance detected in the Time Course dataset (Dark blue: curve fold change > 2.5; Blue: curve fold change > 1.2; White: curve fold change < 1.2 and standard deviation ≤ 0.05; Dotted grey: not detected). Dotted borders indicate non statistically significant oscillating proteins. Yellow borders with the tag indicate proteins also enriched in prometaphase in the Mitotic Exit dataset. D Heatmaps of the top 14 (according to their curve fold change) kinases and components of the ubiquitin-proteasome system (UPS) detected in the Time Course dataset (top panel). Protein changes are coloured according to their abundance (log₂ mean normalised values). In the middle panel, profile plots showing the most significantly oscillating proteins. Number of oscillating (blue) and stable (grey) proteins are shown and proteins validated by Western blot (bottom panel) are in bold. Proteins significantly differentially expressed in the Mitotic Exit dataset are marked with an asterisk. E Venn diagram showing the overlap between oscillating proteins detected in the Time Course proteomics dataset (blue) and oscillating transcripts obtained from Mahdessian et al.¹⁵ (purple). The heatmap further classifies the overlapping proteins according to the cell cycle phase they are peaking in our set and in the transcript level. The colour intensity reflects the percentage of proteins peaking at each time point in our dataset, with darker colours representing higher percentages. F Boxen plot comparing oscillating and stable proteins detected in the Time Course proteomics dataset with protein half-life values obtained from a cycloheximide chase proteomics dataset²⁷. Boxen plots show the median (50th percentile) as a central line. The first box shows the interquartile range, covering 50% of the data (25th to 75th percentiles). Each additional box represents a quantile of half the size of the previous quantile (e.g. box 2 covers 25% of the data (12.5th to 87.5th percentiles), box 3 covers 12.5% of the data (6.25th to 93.75th percentiles) and so on). G Essentiality of oscillating proteins detected in the Time Course proteomics dataset. Scatter plot comparing the curve fold change (oscillation score) and mean gene effect obtained from the Cancer Dependency Map resource²⁸. A lower score indicates a higher probability of gene essentiality, with a score of zero indicating that the gene is not required for cell survival, and a score of −0.5 suggesting essentiality in most cell lines. CCD proteins are shown in blue, and non-CCD proteins are shown in grey. Well-known cell cycle markers are highlighted. Boxen plot shows a significant association between protein oscillation and lower mean gene effect scores (two-sided Mann-Whitney p-value: 2.76 × 10⁻¹⁷). Western blots shown in (D) are representative of at least three independent experiments with similar results. Source data are provided as a Source Data file.