Accuracy evaluation and addition of improved dihedral parameters for the MMFF94s

Abstract

The Platinum dataset of protein-bound ligand conformations was used to benchmark the ability of the MMFF94s force field to generate bioactive conformations by minimization of randomly generated conformers. Torsion angle parameters that generally caused wrong geometries were reparameterized by conducting dihedral scans using ab initio calculations at the MP2 level. This reparameterization resulted in a systematic improvement of generated conformations.

Assessing conformer energies using electronic structure and machine learning methods

“…We expect continued improvement from other methods, particularly multiple efforts to improve classical force fields, [99][100][101][102] inclusion of polarizable atomic charges, [103][104][105][106][107][108][109][110] novel force fields from experimental data, density functional and other quantum methods, [111][112][113][114][115][116] and continued development of approximate semiempirical quantum methods. [37] Currently, we can highly recommend methods at each tier of the accuracy-time tradeoff, particularly the recent GFN2 semiempirical method, the B97-3c density functional approximation, and RI-MP2 for accurate conformer energies.…”

“…We expect continued improvement from other methods, particularly multiple efforts to improve classical force fields, [ 99–102 ] inclusion of polarizable atomic charges, [ 103–110 ] novel force fields from experimental data, density functional and other quantum methods, [ 111–116 ] and continued development of approximate semiempirical quantum methods. [ 37 ]…”

Assessing Conformer Energies using Electronic Structure and Machine Learning Methods

“…We expect continued improvement from other methods, particularly multiple efforts to improve classical force fields, [99][100][101][102] inclusion of polarizable atomic charges, [103][104][105][106][107][108][109][110] novel force fields from experimental data, density functional and other quantum methods, [111][112][113][114][115][116] and continued development of approximate semiempirical quantum methods. 37 At present, we can highly recommend methods at each tier of the accuracy-time tradeoff, particularly the recent GFN2 semiempirical method, the B97-3c density functional approximation, and RI-MP2 for accurate conformer energies.…”

Assessing Conformer Energies using Electronic Structure and Machine Learning Methods

“…We expect continued improvement from other methods, particularly multiple efforts to improve classical force fields, [101], [102], [103], [104] inclusion of polarizable atomic charges, [105], [106], [107], [108], [109], [110], [111], [112] novel force fields from experimental data, density functional and other quantum methods, [113],[114], [115], [116], [117], [118] and continued development of approximate semiempirical quantum methods. [37] At present, we can highly recommend methods at each tier of the accuracy-time tradeoff, particularly the recent GFN2 semiempirical method, the B97-3c density functional approximation, and RI-MP2 for accurate conformer energies.…”