Fig. 4: 10 ns MD screening of selected models in eight groups and 100 ns MD verification of survived models (Model II-1 to II-13, III-1 to III-4, III’-1 to III’-5).

a, b Distribution map of NA domains in the models after 10 ns MD simulation: a Groups I–VI and b Groups III’ and V’. Color denotes \({\chi }_{{{{{{\rm{SAXS}}}}}}}^{2}\) after 10 ns MD simulation. Green, yellow, and orange dots correspond to the structures with\(\,{\chi }_{{{{{{\rm{SAXS}}}}}}}^{2} \, < \, 5.0\),\(\,5.0\le {\chi }_{{{{{{\rm{SAXS}}}}}}}^{2} \, < \, 10.0\) and \(10.0\le {\chi }_{{{{{{\rm{SAXS}}}}}}}^{2}\), respectively. c–f 100 ns MD simulation for 22 models. In each panel, numbers of horizontal axis indicate codes of models corresponding to Table 1. c \({\chi }_{{{{{{\rm{SAXS}}}}}}}^{2}\) (bars) and \({R}_{{{{{{\rm{g}}}}}},{{{{{\rm{SAXS}}}}}}}\) (black circles) for the SAXS profiles averaged over the simulation time (100 ns). d Time evolutions of \({\chi }_{{{{{{\rm{SAXS}}}}}}}^{2}\) in 100 ns MD simulations. e \({\chi }_{{{{{{\rm{SANS}}}}}}}^{2}\) (bars) and \({R}_{{{{{{\rm{g}}}}}},{{{{{\rm{SANS}}}}}}}\) (black circles) for the SANS profiles averaged over the simulation time (100 ns). f Time evolutions of \({\chi }_{{{{{{\rm{SANS}}}}}}}^{2}\) in 100 ns MD simulations. In panels c and e, thick blue lines show the experimental \({R}_{{{{{{\rm{g}}}}}}}\,\)values and thin blue lines express the ranges of their errors.