Browse views: by Year, by Function, by GLF, by Subfunction, by Conference, by Journal

Prediction of fraction unbound in microsomal and hepatocyte incubations – a comparison of methods across industry data sets (by the IQ in silico ADME working group)

Faller, Bernard and Winiwarter, Susanne and Chang, George and Desai, Prashant and Menzel, Carsten and Rieko, Arimoto and Keefer, Christopher and Broccatelli, Fabio (2019) Prediction of fraction unbound in microsomal and hepatocyte incubations – a comparison of methods across industry data sets (by the IQ in silico ADME working group). Molecular pharmaceutics, 16 (9). pp. 4077-4085. ISSN 1543-8392; 1543-8384

Abstract

Fraction unbound in the incubation, fu,inc, is an important parameter to evaluate intrinsic clearance from in vitro CLint measurements in hepatocytes or microsomes. Reliable estimates of fu,inc from structure have the potential to positively impact the screening timelines in drug discovery. Previous works suggest that fu,inc is primarily driven by passive processes and can be described using physico-chemical properties such as lipophilicity and protonation state of the molecule. While models based on these principles proved predictive in relatively small datasets including marketed drugs, their applicability domain has not been extensively explored. The presented work from the in silico ADME discussion group (part of the International Consortium for Innovation through Quality in Pharmaceutical Development, the IQ consortium) investigates the accuracy of these models in larger proprietary datasets considering several thousand compounds across chemical space. Overall, the equations do well for compounds with low lipophilicity, i.e., they correctly predict that fu,inc is in general above 0.5 for compounds with calculated LogP≤3. When applied to lipophilic compounds, the models failed to produce quantitatively accurate predictions of fu,inc with a high risk to underestimate binding properties. Therefore, these models could be used quantitatively for non-lipophilic compounds. Proprietary in-house machine-learning models consistently predict more than 70% of the values within 2-fold of the experimental result regardless of lipophilicity or ionization state. Additionally, data shown indicates that microsomal binding is in general a good proxy for hepatocyte binding.

Item Type: Article
Keywords: microsomal binding, prediction, logP/D, metabolic clearance
Date Deposited: 11 Oct 2019 00:45
Last Modified: 11 Oct 2019 00:45
URI: https://oak.novartis.com/id/eprint/37897

Search

Email Alerts

Register with OAK to receive email alerts for saved searches.