A major focus in medicinal chemistry is the interpretation of structure-activity relationships (SARs) of compounds. The classification of compounds according to different SAR characteristics largely depends on their structural and potency relationships within data sets. Active compounds can participate in different local SAR environments and introduce different degrees of local SAR discontinuity. Such SAR features in large compound data sets are often studied on the basis of activity landscape modeling, which represents a descriptive approach. However, compounds in different local SAR environments have not yet been predicted. To these ends, we have introduced an Emerging Chemical Patterns (ECP)1 approach to systematically predict compounds in different local SAR environments2. The methodology makes it possible to extract key molecular features from specifically active compounds and classify compounds according to different SAR characteristics. A special feature of ECP approach that differs from many other machine learning methods utilized in chemoinformatics is its ability to operate on the basis of small training sets (e.g., as few as 3, 5 or 10 compounds)1. Thus, ECP is an attractive approach for prediction or classification tasks when only limited information is available for model building. For our analysis, compounds from different activity classes were assembled from ChEMBL. The model system developed using ECP was found to accurately predict many compounds in different local SAR environments2. References 1. Auer, J.; Bajorath, J. Emerging Chemical Patterns: A New Methodology for Molecular Classification and Compound Selection. J. Chem. Inf. Model. 2006, 46, 2502−2514. 2. Namasivayam, V.; Gupta-Ostermann, D.; Balfer, J.; Heikamp, K.; Bajorath, J. Prediction of compounds in different local structure-activity relationship using emerging chemical patterns, J Chem Inf Model, 2014, 54(5), 1301-1310.

Expert system for predicting different local structure-activity relationship environments using the concept of emerging chemical patterns

249th National Meeting (2015)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.