Fig. 4: Stability, precision and persistence of functional network structure are markers for cognitive outcome.

A Top: Population vector correlation (PVC) matrices of an exemplary mouse of each outcome group across session pairs within each experimental phase. “Healthy – Stroke” depicts matrices of the last healthy session in relation to the first poststroke session. Dashed lines show reward zone borders. Below: Average PVC curves summarize matrices across corridor ___location offsets. Correlation peaks reflect the periodic structure of the VR corridor. B Y-intercept of PVC curves, which indicates cross-session stability of functional coding, in the three outcome groups (Sham, n = 11; Recovery, n = 5; No-Recovery, n = 4) across session pairs within each experimental phase. Thin lines represent individual mice, thick lines with error bars show mean and standard deviation. Horizontal significance bars mark time effects (group-wise one-way repeated-measures ANOVA), vertical significance bars mark differences between groups in the later phase after stroke ( > 7 days postinjection). C The absolute initial slope of PVC curves, which represents spatial precision in neural ___location coding, with higher values (steeper slopes) indicating higher precision, in the three outcome groups (Sham, n = 11; Recovery, n = 5; No-Recovery, n = 4) across session pairs within each experimental phase. D Change of functional network structure (correlations of the spatially binned activity) over time varies with effect of stroke. Schematic to represent functional network structure before (top) and after (bottom) surgery, with two possible outcomes: functional structure before stroke largely persists (left, solid arrow) or significantly changes (right, dashed arrow). E Example matrices of functional correlations of spatially binned activity on subsequent experimental sessions. In healthy mice, functional structure on a given day (bottom) largely resembles the functional structure on the previous day. After stroke, sham mice and recovery mice exhibit similar functional structure of their spatial activity maps across consecutive days. Instead, No-Recovery mice show very different functional structure of their spatial activity maps even on subsequent sessions. F Cosine similarity of functional correlations of spatially binned activity on different days (computed for the off-diagonal elements of the spatial correlation matrices shown in E). G Mean similarity in the three outcome groups (Sham, n = 11; Recovery, n = 5; No-Recovery, n = 4) across session pairs within each experimental phase. Each data point shows the mean of all similarities in F, when the sessions being compared are both before stroke (Healthy-Healthy), before and after stroke (Healthy-Stroke) and both after stroke (Stroke-Stroke). Boxplots are drawn with the box extending from the 25th to 75th percentiles, with the centre line at the median. Whiskers reach to the minimum and maximum values of the distribution. For A, B data are presented as mean values +/- SD. Group differences in B, C and G were evaluated with two-way repeated measures ANOVA with the Greenhouse-Geisser correction and Tukey-Kramer multiple comparisons tests. Asterisks indicate significances: *p < 0.05, **p < 0.01, ***p < 0.001.