Plan for SMORES v1.0.0

[lukasturcani]

This issue tracks all the relevant tasks needed for the release of SMORES v1.0.0 and provides the context necessary for achieving this. This post is a living doc and will be kept-up-to-date as development on SMORES progresses.

Goals

• Release a Python package which allows users to
  • Calculate sterimol parameters using STREUSEL radii
  • Calculate sterimol parameters from ESP cube files
• Demonstrate the validity of the approach by
  • Generating SMORES parameters for a set of common carbon substituent using
    structures generated with xtb and ESPs generated with Psi4 and comparing these
    to classical sterimol parameters calculated using conventional radii.
  • Generating SMORES parameters for a set of common carbon substituents using
    structures generated with ETKDG using their STREUSEL radii and comparing these
    to classical sterimol parameters calculated using convential radii.
  • Generating SMORES parameters for molecules in catalytic systems described by
    Harper et al.
    • Refactor 6_fit_steric_parameters
    • We need to be able to handle catatlytic systems with 2 substituents
  • Running DBStep on a hollow cube file, which ensures that the voxel grid of the electric field
    surface generated by STREUSEL will be handled correctly.
  • Running DBStep on a nested hollow cube file, which ensures that the voxiel grid of the electric field
    surface generated by STREUSEL will be handled correctly.
  • Verify the fit from the Harper paper given the reported linear fit and the provided experimental ddGs.
  • Implement flour to read/write cube files
  • Fix streusel_cube fit
    • Write streusel_cubes to debug location & view
    • Explore buried streusel_cube sterimol parameters
      • Check if dbStep can do this for us
      • If dbStep is not an option, modify electric field voxels that are sufficiently far from the substituent to zero
• Write an article expanding the utility of the electric field size metric by allowing the calculation of additional steric
  parameters at the electric field surface

Relevant literature

• Multidimensional steric parameters in the analysis of asymmetric catalytic reactions
• Selection of low-dimensional 3D geometric descriptors for accurate enantioselectivity prediction
• Conformational effects on physical-organic descriptors: the case of Sterimol steric parameters
• Practical High-Quality Electrostatic Potential Surfaces for Drug Discovery Using a Graph-Convolutional Deep Neural Network
• ESP-DNN: a graph-convolutional DNN for predicting electrostatic potential surfaces
• Learning atomic multipoles: prediction of the electrostatic potential with equivariant GNNs

[austin-mroz]

• Revert STREUSEL to original method -- rerun cube file generation to eliminate negative values

[lukasturcani]

• Given a distribution of sterimol parameters across conformers, check if the sterimol parameter values reported in Harper are likely

We've found that the values of sterimol parameters we generate are different to those reported in the Harper paper. However they are not that far out and we consider them reasonable

[lukasturcani]

• Check that dbstep uses 0 indexed atoms