Bases de données et modèles QSAR / Databases and QSAR models : Short Courses in Chemoinformatics: Databases and QSAR models

  1. Home
  2. > 2017
  3. > Short Courses in Chemoinformatics: Databases and QSAR models

Informations

Course catalog, page 24

Dates: 15 et 16 Mai 2017.

Code: 1179 Reference: SGI17-0301

Informations and registration :

Sandra GRISINELLI

Tél.: 03 68 85 49 98 (Except on Wednesday)

Fax : 03 68 85 49 29

s.grisinelli unistra.fr

Registration fees :

825 euros These fees include teaching and lunch.

For whom ?

Chemists (Bachelor’s degree or better), technicians having experience in database management, modelisation software, willing to broaden their skills.


Purpose

QSAR modelling (Quantitative Structure Activity Relationship) in Chemoinformatics aims to build predictive statistical models linking the chemical structure of compounds to their physico-chemical or biological properties. These models look for patterns in data stored in chemical databases and use them. Thus, it is important to master the way of storing and using informations contained in molecules. Chemoinformatics approaches are widely used in chemical industry to model physico-chemical properties of molecules and materials, as well as in the pharmaceutical industry to perform virtual screening or to predict pharmacodynamic and pharmacokinetic properties.


Requirements

Basic informatics skills.


Program

Databases management systems in chemistry.
Databases and information sources in chemistry (SciFinder, PubChem, ChEMBL, ChemSpider).
Structural search, sub-structural, superstructural and similarity searches.
In Silico representations of chemical structures : strings (SMILES, SMARTS, INCHI), bit strings (fingerprints), molecular graphs (connectivity table, distance matrix, ...).
Data exchange formats : MOL, RXN, SDF, RDF.

Descriptors, chemical space.
Similarity and diversity of compounds and chemical reactions.
Design of chemical databases.
Data mining methods for chemistry data : MLR, logistic regression, neural networks, SVM, Naïve Bayes, decision trees.
QSAR models for classification and regression : acquisition and validation.
Applicability domain of models.
Ligand-based virtual screening.


Teaching methods

The courses will take place in the Faculté de Chimie, in a computer classroom dedicated to these courses, equipped with 21 PC LINUX, a printer and a video projector.
Lectures will be delivered in English or in French depending on the audience.

Softwares used in the lectures : ChemAxon, MOE, ISIDA, WEKA.


Nature of the course and training approval

This training is an adaptation action and skills development course. A participation certificate will be delivered. At the end of the training, a test will measure the trainees’ satisfaction and achievement of objectives (knowledge, skills, accession, trust) according to levels 1 and 2 of the Kirkpatrick training efficacy assessment template.


Speakers

  • Alexandre Varnek, Professor at Strasbourg University.
  • Gilles Marcou, Senior Lecturer at Strasbourg University.
  • Dragos Horvath, Research Supervisor at the CNRS.

Scientific Leader

M. Gilles MARCOU, Senior Lecturer, Faculté de Chimie.

Email : g.marcou unistra.fr