Integration of Network Biology and SAR Approaches for Early Safety EvaluationCédric Merlot1, Daniel Zuaboni2
|1Drugdesigntech SA, Chemin des Aulx, 14, 1228 Plan les Ouates, Switzerland|
2Leadop Computing, 89 rue du domaine du chateau, 74580 Viry, France
|The aim of this work is to predict adverse drug reactions (ADRs) from the chemical structure through predicted drug-target interaction profiles and biological pathways.|
This multistep strategy has major advantages over current toxicity prediction systems:
- it proposes a mechanism of action,
- the mechanism of action can be translated into short term assays in order to validate or invalidate the prediction in a short timeframe,
- because it is based on more information than only known drug-side effects associations, it is expected to have a greater sensitivity.
Prediction of ADRs from drug-target interaction profiles and the use of pathways has been reported by several groups. However there is currently no integrated system that takes all these methods to provide routinely ADR from the chemical structure on a large scale.
Our methodology starts with the construction of a large number of fragment-based models to predict the drug-target interaction profile: the Predicted Safety Pharmaceutical Profiling (PSPP).