PyMOL - Molecular Graphics System

Links:
https://pymol.org/
https://sourceforge.net/projects/pymol/
https://github.com/schrodinger/pymol-open-source
https://www.schrodinger.com/platform/
https://www.schrodinger.com/materials-science/use-cases/specialty-chemicals/

List of other Cheminformatics Libraries-Platforms (a lot based on RDKit):
https://github.com/hsiaoyi0504/awesome-cheminformatics



Videos can be viewed here:  https://www.schrodinger.com/materials-science/solutions/catalysis-and-reactivity/

Accelerate the optimization and discovery of high-performance chemicals with digital chemistry

With Schrödinger’s digital chemistry platform, you have the tools to develop more reliable and cost-efficient specialty chemicals with excellent performance. Leverage molecular simulation and machine learning for in silico design of novel monomers and additives that are used in plastics and rubbers, industrial coatings and paints, adhesives and personal care products, as well as catalysts used in the production of such chemicals.

https://jcheminf.biomedcentral.com/articles/10.1186/s13321-020-00479-8
Molecular representation

Graph-based methods are capable of learning molecular representations by operating the convolutions on the encoded molecular graphs directly. In the graph representation for a molecule, the connectivity relation between atoms is represented by a graph G = (V, E). Here, the nodes V are represented by the node feature vector Xv consisting of a series of atomic features and the edges E are represented by the edge feature vector Ekm consisting of a series of bond features, where the subscript km indicates that atoms k and m are bonded. Followed by previous studies [27], almost all the easily accessible atom/bond-level features were exhausted to comprehensively squeeze chemical information into molecular graph for graph-based models, where include nine kinds of atomic features (i.e., atom symbol, atom degree, formal charge, radical electrons, hybridization, aromaticity, hydrogens, chirality and chirality type) and four kinds of bond features (i.e., bond type, conjugation, ring, and stereo). Most of them were encoded into a molecular graph in a one-hot manner and subsequently the encoded molecular graph was used as the input. The more details about the molecular representations for graph-based models are available in the publication [27].