Explore the concept of chemography in chemical space analysis through this 36-minute lecture by Professor Alexandre Varnek. Delve into topics such as encoding chemical structures, data visualization techniques, dimensionality reduction methods, and Generative Topographic Mapping (GTM). Learn about density landscapes, chemical analysis with density maps, and the application of GTM in virtual screening. Discover the comparative analysis of large chemical libraries, hierarchical GTM navigation, and the generation and analysis of DNA-encoded libraries. Examine the use of autoencoders in SMILES reconstruction and chemical space exploration. Gain insights into AI-driven design of new chemical transformations and Suzuki-like reactions. Benefit from Professor Varnek's extensive experience in chemoinformatics and molecular modeling as he shares his expertise in this comprehensive overview of chemography and its applications in chemical space analysis.
Chemography Concept in Chemical Space Analysis - 2022
School of Chemoinformatics in Latin America via YouTube
Overview
Syllabus
Intro
Encoding chemical structures by molecular descrip
Data visualization: dimensionality reduction proble
Dimensionality reduction methods
Generative Topographic Mapping: algorithm
Probabilistic nature of GTM
GTM Density landscape
Density landscapes
Chemical analysis with density maps
Encoding a library by a vector using GTM landscape
Tunable ISIDA fragment descriptors
ChemSpace Atlas tool
Universal maps: application to virtual screening
Comparative analysis of (ultra)large chemical librari
Commercial vs Biologically relevant data
GTM class landscape for libraries comparison
Hierarchical GTM navigation of the chemical space
Generation and analysis of general-purpose DELS
Selection of DEL the best covering a reference library (ChEMBL) c
Linear vs chemographic property distribution
Autoencoder performing SMILES reconstruction
AutoEncoder: sampling using a seed vector
AutoEncoder chemical space: choice of a seed vect
Al-driven design of new chemical transformations
Al-driven design of new Suzuki-like reactions
Taught by
School of Chemoinformatics in Latin America
