Fig. 3

Regulation of variable complex members. a Co-regulation of protein abundance change and transcript expression was investigated for the reprogramming dataset using published data derived from the same time course experiment [12, 22]. Transcriptional regulation was inferred when both protein and mRNA were significantly regulated with a consistent fold change at the same time point of reprogramming. Thirty-eight percent of the investigated protein changes (223 in total) co-occur with changes in mRNA level (orange dots), indicating a transcriptional regulation of complex member abundance. For about half of these cases, the change in transcript level is anti-correlated with the expression of at least one miRNA targeting the same mRNA [22]. The regulation of these complex members might therefore be mediated by a miRNA-based mechanism. b Changes of complex members that are not transcriptionally regulated often occur via a decrease in protein abundance. Violin plots show the distribution of protein fold changes detected for variable complex members. Proteins are grouped according to whether transcriptional regulation could be inferred from mRNA expression data, as shown in a. c Variable complex members tend to be more phosphorylated than stable one. Phosphosites were derived from PTMcode [60]. d Violin plots for core and variable interfaces show the distribution of desolvation energies per square angstrom of apolar buried surface; stable interfaces have more favorable hydrophobic interactions (Wilcoxon rank sum test, p value = 8.8E-6). See Methods for details. e Core interface of two exosome components have hydrophobic residues that are evenly distributed in the interface core and surrounded by hydrophilic residues leading to favorable hydrophobic interaction. f Variable interface of two COPII components: here, hydrophobic and hydrophilic residues are mixed in the interface, leading to unfavorable hydrophobic interactions. Interfaces shown in e and f include residues within 6 Å of inter-atomic distance