Fig. 3

Changes in glucokinase activity explained by decreased abundance and conformational shifts. A Abundance and activity scores of 9019 missense, nonsense, and synonymous glucokinase variants shown as a 2D histogram. The thresholds for low abundance (0.58) and low activity (0.66) are indicated as red dashed lines. The number and percentage of variants falling within each quadrant are reported. B The median abundance and activity of variants at each position of the glucokinase sequence is shown as a line plot. The regions forming the hinge region (gray) and the large (light blue) and small (light orange) domains are represented as a bar at the bottom. C Plots zooming in on regions 145–190 (left) and 330–400 (right) from panel B. D Barcode plots showing the median abundance score, predicted change in protein thermodynamic stability (\(\Delta \Delta G\), kcal/mol) using the closed active state or the super-open inactive state, and the difference \((\Delta (\Delta \Delta G))\) between the two \(\Delta \Delta G\) predictions. For the bottom plot, red indicates that variants at these positions are predicted to destabilize the closed state more, while at blue positions variants are predicted to destabilize the super-open state more. PDBs: 1V4S (closed) and 1V4T (super-open). The \(\Delta \Delta G\) data were obtained from [22] except for the 157–179 region in the super-open state. For all panels, the data on glucokinase variant activities were obtained from [22]