Fig. 6. Bootstrapped Bayesian networks indicate that SARS-CoV-2 infection changes organelle cross-regulation.

A–C Averaged Bayesian networks representing the dependencies (equivalence classes) between mitochondria, nuclei, nucleoli, lipid droplets, and cytosol dry mass states and the syncytial state of A non-infected cells, B infected cells, and C Syncytin-1-expressing cells over 10,000 randomly resampled datasets. All edges at each iteration were validated by two-sided chi-squared, g-squared, log-likelihood, and Cressie-Read pairwise tests, all displaying a p-value < 10⁻⁶ over discretized and non-discretized data (see Supplementary Tables 1–4 for two-sided chi-squared test details) (# of independent time-lapse acquisitions for each condition ≥ 4).