To know the main approaches to the computer representation of molecular structures and chemical reactions. To know how to represent features of molecular structures by molecular descriptors. To know the fundamentals of QSAR/QSPR methodologies and its application with multilinear regressions, decision trees and neural networks.
João Montargil Aires de Sousa
Weekly - 2
Total - 28
1. Chemoinformatics - a Textbook, Gasteiger, J. Engel, T., Eds.; Wiley-VCH: Weinheim, 2003.
2. Leach, A. R.; Gillet, V. J. An Introduction to Chemoinformatics, 2ª ed.; Springer: Dordrecht, 2007.
3. Short texts, tutorials, solved problems, made available via the Moodle system.
4. Applied Chemoinformatics: Achievements and Future Opportunities, Engel, T. Gasteiger, J. , Eds.; Wiley-VCH: Weinheim, 2018.
This course uses Team-Based Learning, TBL, http://www.teambasedlearning.org .
The unit is organised in 4 modules. Before each module, students are provided with the learning material and a list of specific objectives. Before the first class of each module, each student must answer an individual test (Readiness Assurance Test). The same test is answered by teams in class, followed by a mini-lecture to solve the test, discuss doubts and reinforce the most difficult points.
In the other classes of the module, teams are challenged with application activities.
Class evaluation: 60%, Final test/exam: 40%
Minimum marks: 8/20 in class, 10/20 in final test/exam.
Mark for class activities=average of individual TGPitests (50%) and team results (25% tests, 75% activities). Mark corrected by peer evaluation (team mark x points received by colleagues/100).
There is no second attept to improve marks for class activities in the same year. Students have access to exams to improve the TP mark (the final test).
1. Representation of molecular structures: linear notations, molecular graphs, connectivity tables, structural keys, hashed fingerprints and hash codes.
2. Representation of chemical reactions.
3. Molecular descriptors.
4. Data analysis and property prediction (QSPR/QSAR): multilinear regressions, decision trees and neural networks.