Automated QSPR

Cartmell, J., Enoch, S., Krstajic, D., Leahy, D. E. Automated QSPR through... 

Cartmell J, Enoch S, Krstajic D, Leahy DE. Automated QSPR through competitive workflow. J Comput Aided Mol Des 2005 19 821-33. 

The process described in the preceding paragraphs has seen widespread use. This is partly because it has been automated very well in the more sophisticated QSPR programs. 

Horvath D. (2001b) ComPharm—Automated comparative analysis of phar-macophoric patterns and derived QSAR approaches, novel tools in high throughput drug discovery. A proof-of-concept study applied to farnesyl protein transferase inhibitor design. In M Diudea (ed), QSPR/QSAR Studies by Molecular Descriptors, pp. 395-439, Nova Science Publishers, New York, USA. 

The structure of each compound is stored as a connection table. A molecular models is generated for each stored structure using molecular mechanics model building such as MM2, the semiempirical method MOPAC 6.0, or specialized methods such as a recently developed extended Hiickel method. Three-dimensional structures can also be generated directly from their connection tables by structure generators (see Three-dimensional Structure Generation Automation) such as concord or CORINA. Some approaches to QSPR use only descriptors derived from the topological representation of the molecular structures, and in this case the development of three-dimensional molecular models is not necessary. 

Semichem s partnership with the University of Florida to produce the QSAi QSPR program CODESSA has further enhanced the usability and general applicability of Semichem s software. CODESSA ties AMPAC results (thermodynamic, electrostatic, geometric, and quantum mechanical) to experimental results, making real-life application easier and more intuitive. Together AMPAC and CODESSA can often predict actual product characteristics and properties. Semichem has also added a PREDICT module, which automates the execution of both AMPAC and CODESSA, with the objective of generating predictions for molecules using an already derived correlation. 

Computational techniques intended to automate generation and mining of virtual libraries of compounds have successfully been developed and used in recent years for molecular and materials discovery and optimization. This activity has been extensively used in the field of drug discovery where quantitative structure property relationships and quantitative structure activity relationships (QSPR/QSAR) have been used to build correlations between structural molecular features and the experimentally measured properties or activities of the molecules. QSPR/QSAR approaches have been reported in the literature to predict many physicochemical properties, such as vapor pres-... 

---