Fig 4. Matrix feedback generates diverse patterning.
(A) Matrix patterns produced from varying noise and cell-matrix feedback, cell-cell guidance fixed at wc = 0.03. Simulations are of 800 cells (40% confluence) over a time-course of seven days. From left to right noise η = (0, 0.07, 0.14, 0.21), from top to bottom wm = (0, 0.04, 0.12, 0.2). Scale bar represents 100μm. (B) Four representative matrix patterns from A are characterized with a starplot according to the metrics of long-range alignment (LRA), short-range alignment (SRA), percentage of high density matrix (HDM), curvature (Curv) and fractal dimension (Frac). (C) PCA of simulations exploring parameter space, reducing metric-space down into two prinicpal components explains 85% of variance and shows the areas in metric-space that can be reached with matrix feedback.