Related presentations, posters, and scientific talks from this event have been posted here for your reference. Please click the associated link to download.
| Title | Author | Link |
|---|---|---|
| In-silico techniques in lead optimization—perspectives for rational use | Pranas Japertas, Andrius Sazonovas, Kiril Lanevskij | Download Poster |
Designing lead molecules with appropriate physicochemical properties is the ideal starting point in drug discovery, and is the reason for extensive in silico and in vitro compound screening. The structure design application in ACD/Percepta was designed for synthetic chemists, that aids chemical structure optimization based on pre-defined physicochemical, ADME, and toxicity requirements. Customize your level of prediction detail, use various tools to filter and sort results, and apply experimental data to improve prediction accuracy. Meet us at the Medicinal Chemistry Symposium to discuss how this and other solutions can help accelerate your projects.
Title: In-silico techniques in lead optimization—perspectives for rational use
Authors: Pranas Japertas, Andrius Sazonovas, Kiril Lanevskij
Date: May 20–23, 2012
Abstract: View Abstract
Despite constant advances in the QSAR field, computational approaches in general still fail to meet the high expectations associated with their use. In this work we attempt to re-assess the role of various in-silico tools in building an effective lead optimization strategy.
The major issue with most in-silico models available up to date is that they are 'statistics driven' and have been developed with the only pursuit of the best possible fit in mind. At the same time it is usually not taken into account that many of the analyzed properties are governed by the most basic physicochemical characteristics of the compounds, such as ionization, lipophilicity, molecular size, or hydrogen bonding potential. In this work, we demonstrate that this set of parameters suffices for devising accurate models of intestinal absorption and passive permeation across the blood-brain barrier. Moreover general physicochemical trends are observed even for the properties involving protein-ligand interactions (P-gp, hERG). The key advantage of simple physicochemical models is their interpretability. Not only they produce an estimate of the property value for a compound, but also provide a mechanistical insight that has the potential to guide lead optimization in the desired direction. In our opinion, a rational strategy of lead optimization should aim at achieving balanced physicochemical profiles of candidate compounds that would translate into favorable ADME properties.
Another aspect covered in this work is the phenomenon of 'activity cliffs'. It manifests as local anomalies when small structural changes lead to disproportionally large changes in observed effects. The issue could be potentially resolved by employing advanced modeling techniques, such as pairwise QSAR. Additional inclusion of target affinity data in the simple ADME/Tox profiling analysis is introduced next as a natural extension of all above concepts. Available potency data for the considered compound series could be utilized with the help of automated Hansch and Free-Wilson type analysis (Auto-SAR approach). Analog generation combined with automated modeling has the potential to suggest most promising candidates.
The factors that will impact the commercial fate of a particular analogue are not limited by its potency and ADME/Tox profile. There are also a number of aspects that are not associated with compound’s suitability to be used as a drug, namely synthetic feasibility and patentability prospects. The possibilities of their assessment using cheminformatics approaches (such as comparison of substructures of different sizes in the analyzed compounds with the ones in available chemistry) are also overviewed in this work.
