Fig. 5: Statistical analysis and simulations.

We used analysis of Variance (ANOVA) test to compare the small-world-ness of networks formed on different surfaces. Multiple-comparison post hoc Bonferroni test (a series of t-tests performed on each pair of groups corrected by the number of groups) indicates which samples means are significantly different from the control, i.e. random neuronal-networks cultured on rigid surfaces with \({{{{{\rm{SW}}}}}}=1\). In the diagram, sample-means that are different at some significance level \(\alpha\), are marked by a bar. If \(\alpha\) is less than \(0.05\), it is flagged with \(1\) star (*). If \(\alpha\) is less than \(0.01\), it is flagged with \(2\) stars (**) (a). We used information theory to estimate the amount of information transported in networks of neuronal cells. We built connected graphs from fluorescence images of cells on the substrates (b) and examined how an initial disturbance propagates in those networks - resulting space and time patterns of signals were used to estimate the information processed over time in each node (c). Results of this theoretical analysis: total information \(I\) elaborated in neuronal cell graphs cultured on soft PDMS surfaces, as a function of surface elasticity (d). Data in Fig. 5a and d are represented by mean ± standard deviation. The sample size for data reported in Fig. 5d is 10.