Fig. 5: Genes influencing EMT intermediate state dynamics are identified via consensus analysis of expression, regulation, and parameter inference.

a Annotation of predicted dynamic intermediate state genes by cellular location. b Annotation of predicted dynamic intermediate state genes by EMT marker gene sources. c SFN is a predicted dynamic intermediate state gene, with its differential expression in multiple EMT intermediate states and significant correlations with model parameters k₁ and k₂. d NRG1 is another predicted dynamic intermediate state gene, with its differential expression in multiple EMT intermediate states, significant negative correlation with model parameter k₂, and differential regulation in EMT intermediate states via RNA velocity.