In silico prediction of absorption, distribution, metabolism and elimination (ADME)

It is estimated that approximately one half of newly proposed drugs fail during the clinical phase because of their high toxicity and/or poor Absorption, Distribution, Metabolism and Excretion (ADME) characteristics. The progress observed in recent years from the use of combinatorial chemistry and high-throughput screening resulted in a large number of chemicals being screened and tested for specific biological activities. This resulted in an increased need to evaluate ADME properties, toxicity and potency of large sets of compounds. As expected, the demand for rapid and predictive ADME models rose significantly. In this work, an attempt was made to develop computational models capable of predicting each of the ADME properties.; The ability of a drug to cross the intestinal membrane, to be distributed throughout the blood stream and tissues, to avoid metabolism (if metabolism is not desired), and to be eliminated from the body depends on physicochemical properties. The Multiple Computer Automated-Structure Evaluation (MCASE) Program, which was used for the development of the majority of the computational models presented in this study; is known for its ability to find meaningful relationships (when they exist) between the physicochemical properties of the compounds of interest and their biological and ADME characteristics.; Human intestinal absorption potential was evaluated by a quantitative structure-activity relationship (QSAR) that included structural descriptors derived from the chemical structures of a data set containing 417 drugs. The ability of chemicals to bind to different plasma proteins as well as the volume of fluid in which they distribute was estimated by simple MCASE models that include as descriptors structural moieties deemed biophores. Models to estimate the extent to which a drug is metabolized and excreted as unchanged/changed were also developed and for each of them an attempt was made to explain the involvement of the biophores in the studied process.; The availability of reliable and fast models to predict ADME properties such as those presented in this work could potentially speed up the process of identifying compounds with improved properties, ultimately making the entire drug discovery process shorter and more cost effective.