Investigating the\nOrigin of Automatic Rhodopsin Modeling\nOutliers Using the Microbial Gloeobacter Rhodopsin as Testbed

The automatic rhodopsin modeling (ARM) approach is a\ncomputational\nworkflow devised for the automatic buildup of hybrid quantum mechanics/molecular\nmechanics (QM/MM) models of wild-type rhodopsins and mutants, with\nthe purpose of establishing trends in their photophysical and photochemical\nproperties. Despite the success of ARM in accurately describing the\nvisible light absorption maxima of many rhodopsins, for a few cases,\ncalled outliers, it might lead to large deviations with respect to\nexperiments. Applying ARM toGloeobacter rhodopsin (GR), a microbial rhodopsin with important applications\nin optogenetics, we analyze the origin of such outliers in the absorption\nenergies obtained for GR wild-type and mutants at neutral pH, with\na total root-mean-square deviation (RMSD) of 0.42 eV with respect\nto the experimental GR excitation energies. Having discussed the importance\nand the uncertainty of one particular amino-acid p<i>K</i><sub>a</sub>, namely histidine at position 87, we propose and test\nseveral modifications to the standard ARM protocol: (i) improved p<i>K</i><sub>a</sub> predictions along with the consideration of\nseveral protonation microstates, (ii) attenuation of the opsin electrostatic\npotential at short-range, (iii) substitution of the state-average\ncomplete active space (CAS) electronic structure method by its state-specific\napproach, and (iv) complete replacement of CAS with mixed-reference\nspin-flip time-dependent density functional theory (MRSF-TDDFT). The\nbest RMSD result we obtain is 0.2 eV combining the protonation\nof H87 and using MRSF/CAMH-B3LYP.