Fig. 7: Mitigation of perturbations via feedback control.

a The CL and OL systems introduced in Fig. 6 were tested against two perturbations: (i) indirect transcriptional inhibition via loading of transcriptional resources by Gal4-VPR and (ii) direct post-transcriptional knockdown by miR-FF4. The kinase, perturbations, and controllers were each poly-transfected in separate DNA-lipid complexes in order to measure 2D dose-response of the OL and CL systems to the kinase and perturbations (see Supplementary Figures 19–22 for details). b Dose–responses of OL and CL systems highlighted in the following panels. The Fluc2 and Fluc2/3 OL variants were chosen since they have nearly identical output levels compared to the CL with and without DDd, respectively, in the absence of kinase. Dose–responses and detailed comparisons among all OL and CL variants are provided in Supplementary Figures 25–28. c Fold-changes (Fold-Δs) in output expression in response to miR-FF4 (top row) and Gal4-VPR (bottom row) perturbations. Each column represents an increasing amount of kinase input from left to right. The dashed lines indicate no fold-change (ideal). d Direct comparison of fold-changes to perturbations between OL and CL variants across kinase dosages. The data represents the maximum dosage of miR-FF4 (miR Marker ≈ 10^7.75 MEFLs) and dosage of Gal4-VPR with a comparable level of knockdown to the OL (Gal4 Marker ≈ 10^6.25 MEFLs). e Robustness scores (100% − % deviation due to perturbations) for all OL and CL variants across each kinase input level at the same dosages of miR-FF4 and Gal4-VPR as highlighted in d. Nominal outputs indicate the level of output in the absence of any perturbations. The individual points are drawn from all experimental repeats. All data were measured by flow cytometry at 48 hours post transfection in HEK-293FT cells. All error bars represent the mean ± standard deviation of measurements from three experimental repeats. Source data are provided as a Source Data file.