Global Analysis of Pharmacophoric Space
Lead Research Organisation:
University of Cambridge
Department Name: Chemistry
Abstract
Rationale
It is well known that many drug molecules bind to, and are active against, multiple protein targets. Recent studies suggest that this promiscuous behaviour is a consequence of the finite number of protein pocket shapes. Furthermore, in these ligandable pockets there are special highly energetic local regions (so called hot spots) that drive the small molecule-protein binding. These regions are thought to be conserved across protein families and conformational ensembles. If the shapes and chemical properties of these hot regions in proteins can be identified, and encoded as pharmacophores, then we will be in position to estimate the number of small molecule ligands needed to cover them. These pharmacophores can also potentially be used to design or select a set of compounds for this purpose. The high success rate of fragment-based screening is a validation of this approach. We aim to discover how to achieve this success with larger compounds, ones which bind with sufficient affinity to have a functional effect.
Project Summary
This project will overlap with and take forward the hotspot prediction work of Chris Radoux, a PhD student sponsored by BBSRC-UCB and hosted by the CCDC. The plan is that Chris will develop a method to automatically derive pharmacophores from his hotspot predictions. This project will be to run use this method on all pockets in PDB and analyse the output. Finally these pharmacophores will be used to design and/or select compounds that fit as many of these pharmacophores as possible. Of course, the number of compounds needed to match all pharmacophores will be a function of their complexity. Part of this project will be to calculate what this relationship is.
It is well known that many drug molecules bind to, and are active against, multiple protein targets. Recent studies suggest that this promiscuous behaviour is a consequence of the finite number of protein pocket shapes. Furthermore, in these ligandable pockets there are special highly energetic local regions (so called hot spots) that drive the small molecule-protein binding. These regions are thought to be conserved across protein families and conformational ensembles. If the shapes and chemical properties of these hot regions in proteins can be identified, and encoded as pharmacophores, then we will be in position to estimate the number of small molecule ligands needed to cover them. These pharmacophores can also potentially be used to design or select a set of compounds for this purpose. The high success rate of fragment-based screening is a validation of this approach. We aim to discover how to achieve this success with larger compounds, ones which bind with sufficient affinity to have a functional effect.
Project Summary
This project will overlap with and take forward the hotspot prediction work of Chris Radoux, a PhD student sponsored by BBSRC-UCB and hosted by the CCDC. The plan is that Chris will develop a method to automatically derive pharmacophores from his hotspot predictions. This project will be to run use this method on all pockets in PDB and analyse the output. Finally these pharmacophores will be used to design and/or select compounds that fit as many of these pharmacophores as possible. Of course, the number of compounds needed to match all pharmacophores will be a function of their complexity. Part of this project will be to calculate what this relationship is.
Organisations
People |
ORCID iD |
| Peter Curran (Student) |
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| BB/P50466X/1 | 30/09/2016 | 29/03/2021 | |||
| 1849980 | Studentship | BB/P50466X/1 | 30/09/2016 | 30/08/2021 | Peter Curran |
| Description | We have developed a Python API enabling flexible usage of the Fragment Hotspot Maps algorithm. Fragment Hotspot Maps identifies the locations and the required intermolecular interactions of small molecule binding hotspots, we have extended the capabilities to detect charged interactions. The API also enables the selected pharmacophoric "fields" to be used in Virtual Screening. We support 3 methods: Docking (with GOLD), Pharmacophore search (with CrossMiner and Pharmit), and Field-based Screening (Unpublished tool). |
| Exploitation Route | The API is actively being used by collaborations and industrial partners: - To investigate selectivity profiles - To investigate target tractability - For virtual screening |
| Sectors | Chemicals Pharmaceuticals and Medical Biotechnology |
| Title | Fragment Hotspot Maps Python API |
| Description | The Hotspots API is the Python package for the Fragment Hotspot Maps project, a knowledge-based method for determining small molecule binding "hotspots". For more information on this method: Radoux, C.J. et. al., Identifying the Interactions that Determine Fragment Binding at Protein Hotspots J. Med. Chem. 2016, 59 (9), 4314-4325 |
| Type Of Technology | Software |
| Year Produced | 2019 |
| Open Source License? | Yes |
| Impact | This product has enabled collaboration and is supporting the application of Fragment Hotspot Maps analysis to structure based drug discovery methods |
| URL | https://github.com/prcurran/hotspots |
| Description | Hotspots User Group Meetings |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Other audiences |
| Results and Impact | Hotspots UGM occurs at 6 monthly intervals and is an opportunity for users of the Fragment Hotspot Maps algorithm to share research outcomes and ideas for the future of the project. |
| Year(s) Of Engagement Activity | 2019 |