具体问题如下：
1. Map DCCM (in Bio3D) for RMSF data of each variant, to establish whether the flagged residues above for RMSF are involved in any key networks that affect stability and/or the active site.
2. Is there a way to combine the analysis of RMSF and RMSD, eg. In a ML or other approach? RMSD and activity correlate reasonably well, whereas RMSF and stability correlate well at certain residues. A combined RMSD and RMSF correlation may work better?