Identification of factors and mechanisms underlying changes in cell phenotype profiles at the G0/G1 transition. ARPE-19 cells were serum starved to arrest in G0 and induce ciliogenesis. Cell phenotype profiles were then analyzed 18 h after serum re-addition under various experimental conditions. (A) Time course of serum starvation and serum re-addition. Five experimental conditions were established by combining siRNA (with NC as the nontargeting control) and drug treatments (DMSO or SP), as detailed in the table on the right. EdU was added at the time of serum re-addition only in experiments designed to measure EdU incorporation. (B) Monitoring cell cycle progression and ciliation under each condition. Raincloud plots (Allen et al., 2021) show the distributions of the mean nuclear EdU (mean_EdU) and Ki-67 (mean_ki67) fluorescence intensities after background subtraction, alongside the percentage of ciliated cells (calculated as the ratio of cilia to nuclei). Representative images are shown in Fig. S3 D. (C) Representative cropped images of cells used in the cell phenotype profiling. Additional images are provided in Fig. S3 E. Arrows indicate cilia. AcTub, acetylated tubulin. Scale bar, 20 µm. (D) UMAP visualization using features from Fig. S3 C. Data for each experimental condition are highlighted separately for clarity. Two major clusters emerged, representing a predominantly non-ciliated group and a smaller ciliated group. (E) Analysis of notable features. UMAP and violin plots illustrate the distributions of several key parameters across experimental conditions: mean_nuc (mean DAPI/DNA fluorescence), area_Golgi (Golgi area), mean_Golgi (mean GRASP65/Golgi fluorescence intensity), peaks_num_Golgi (number of GRASP65 fluorescence peaks), peaks_xy_std_Golgi (SD of fluorescence peak coordinates), and dist_nuc_Golgi (distance between the nucleus and Golgi). Full data are available in dataset 6.