25) and for all further analysis the wave velocities of both strains were combined. Availability of supporting data The data sets supporting the results of this article are available in the 3TU.Datacentrum repository [56], [doi:10.4121/uuid:f5603abf-bf15-4732-84c0-a413ce7d12d3], [http://dx.doi.org/10.4121/uuid:f5603abf-bf15-4732-84c0-a413ce7d12d3]. Acknowledgments We thank Martin Ackermann, Robert H. Austin, Jean-Baptiste

Boulé, Cees Dekker, Alex Hall, Rutger Hermsen and Pieter Schoustra for valuable comments and discussion CYC202 in vitro and Orsolya Haja for measuring the bulk growth curves. The project described was supported by Grant Number U54CA143803 from the National Cancer Institute. The content is solely the responsibility of the authors and does

not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health. P.G. was supported by the “Lendület” program of the Hungarian Academy of Sciences. Electronic supplementary material Additional file 1: Growth curves of strains JEK1036 and JEK1037 in bulk conditions. Growth curves are shown for strains JEK1036 (in green) and JEK1037 (in red), for each strain 3 independent cultures were grown in 200 ml LB in 500 ml flasks at 30°C. For each sample the OD600 was measured in triplicate and their average value was check details used. Error bars indicate sem. The inset shows the growth curve using linear y-scale for the first 15 hours. (PDF 104 KB) Additional file 2: Overview of all devices with separate inlets (type 1).

(A) Each kymograph shows the average occupancy per patch in a single habitat. Kymographs for the five parallel habitats in a single device are shown next to each other. Note that all habitats on the same device are inoculated from the same culture set. (B) The device-wide averages of the occupancies of strains JEK1037 (R red) and JEK1036 (G green) and the red fraction (f r black) are shown as function of time. Dashed lines indicate mean ± sem. The red fraction (f r ) is calculated for each habitat as f r  = r/(r + g), where r and g are the habitat-wide average G protein-coupled receptor kinase occupancies of strains JEK1037 (red) and JEK1036 (green) respectively. Habitats where one (or both) of the strains failed to enter (e.g. when there is a constriction in one of the inlet channels) were excluded from the analysis and are shown as grey panels in this figure. (PDF 443 KB) Additional file 3: Overview of all devices with a single inlet (type 2). (A) Each kymograph shows the average occupancy per patch in a single habitat. Kymographs for the five parallel habitats in a single device are shown next to each other. Note that all habitats on the same device are inoculated from the same culture set. (B) The device-wide averages of the occupancies of strains JEK1037 (R, red) and JEK1036 (G, green) and the red fraction (f r black) are shown as function of time. Dashed lines indicate mean ± sem.