Exploration of the Dynamic Properties of Protein Complexes Predicted from Spatially Constrained Protein-Protein Interaction Networks
A: The axis along the path of directed motion is expected to have the highest variance. Using the covariance matrix from the fitted Gaussian for the Protein of Interest (POI), the eigenvector (EV) with the largest absolute eigenvalue (EV1) forms the axis along which diffusion is occurring. B: Diffusion is occurring either in the direction of the EV or in the opposite direction. The long tail of the distribution determines the direction of diffusion, which can be calculated from the skew of the distribution. C: Activation of the Arp2/3 complex by Las17. Arp2 and Arp3 move towards each other, and Las17 moves towards (and binds) to Arp2 and Arp3. Sub-region 1 contains conformations that are consistent with Arp2/3 activation. D: Visualization of sub-region 1 Arp2, Arp3 and Las17 conformations. E: Visualization of Arp2, Arp3 and Las17 conformations in sub-regions 2-5. F: Directed diffusion amongst Arp2, Arp3 and Las17. Arrow length represents diffusiveness (EV1) relative to the center of mass (COM; circles) of the distribution of conformations for each protein. Numerical value is the skew, or direction of diffusion. Higher values correspond to increasing bias. In sub-region 1, Arp2 and Arp2 move towards each other and Las17 moves towards Arp2 and Arp3.