Autonomous Discovery of Functional Small Molecules
Lead Research Organisation:
University of Leeds
Department Name: Sch of Chemistry
Abstract
Small molecule drugs continue to dominate Man's ability to treat disease. However, the pharmaceutical industry faces unprecedented challenges on multiple fronts. The sector is dogged by a crippling failure rate in drug discovery (over 95%), and a current cost/launch of ~£1.8bn. Around 45% of the costs fall early in drug discovery because relatively few campaigns proceed to the launch of a therapeutic product. Greater innovation is urgently required to address the sector's grand challenge of improving productivity.
The binding of a drug to its target protein is broadly analogous to the fitting of a key into a lock. A major challenge is to optimise the structure of the drug so that it complements its target protein perfectly. Current practices rely on the synthesis of many molecules that are then individually purified and then tested for improved biological function. Many cycles are required to discover a molecule with optimal properties, and much effort is expended in the synthesis and purification of poorly active compounds. Optimisation is thus lengthy, laborious and resource-intensive.
My fellowship will enable me to develop an autonomous platform for the discovery of functional molecules. The approach will exploit chemistry that enables the function of many different molecules to be explored in parallel, in order to determine features are important for biological activity. The whole discovery approach will be automated, with the functions of the products in one round directly informing the design of reactions in the next round. Unlike most medicinal chemistry approaches, products will only be purified and identified when exciting improved biological activity is detected. The approach is thus remarkably efficient because resources are focused overtly on only those reactions that yield molecules with improved function. The resulting molecules may serve as highly innovative starting points for drug discovery programmes.
I will show that my novel approach can support complementary strategies for functional molecule discovery. Crucially, I will demonstrate that it can enable the discovery of ligands - "keys" - for a range of increasingly challenging, medicinally-relevant, target protein. This activity will ensure that my novel approach is fully appreciated by scientists engaged in small molecule discovery. To ensure alignment with future discovery needs, I will collaborate with two major pharmaceutical companies and engage extensively with a wide range of drug discovery organisations. At the end of the project, I will define a mechanism that will ensure that my approach is accessible to end-users in discovery-based industries. Thus, I will ensure that my novel platform can facilitate the discovery of future drug candidates.
The binding of a drug to its target protein is broadly analogous to the fitting of a key into a lock. A major challenge is to optimise the structure of the drug so that it complements its target protein perfectly. Current practices rely on the synthesis of many molecules that are then individually purified and then tested for improved biological function. Many cycles are required to discover a molecule with optimal properties, and much effort is expended in the synthesis and purification of poorly active compounds. Optimisation is thus lengthy, laborious and resource-intensive.
My fellowship will enable me to develop an autonomous platform for the discovery of functional molecules. The approach will exploit chemistry that enables the function of many different molecules to be explored in parallel, in order to determine features are important for biological activity. The whole discovery approach will be automated, with the functions of the products in one round directly informing the design of reactions in the next round. Unlike most medicinal chemistry approaches, products will only be purified and identified when exciting improved biological activity is detected. The approach is thus remarkably efficient because resources are focused overtly on only those reactions that yield molecules with improved function. The resulting molecules may serve as highly innovative starting points for drug discovery programmes.
I will show that my novel approach can support complementary strategies for functional molecule discovery. Crucially, I will demonstrate that it can enable the discovery of ligands - "keys" - for a range of increasingly challenging, medicinally-relevant, target protein. This activity will ensure that my novel approach is fully appreciated by scientists engaged in small molecule discovery. To ensure alignment with future discovery needs, I will collaborate with two major pharmaceutical companies and engage extensively with a wide range of drug discovery organisations. At the end of the project, I will define a mechanism that will ensure that my approach is accessible to end-users in discovery-based industries. Thus, I will ensure that my novel platform can facilitate the discovery of future drug candidates.
Planned Impact
The non-academic beneficiaries of the research are:
1. Companies engaged in molecular discovery
The discovery of small molecule drugs is undertaken within the pharmaceutical sector (comprising large pharmaceutical companies) and the medical biotechnology sector (that includes smaller drug discovery companies). These companies face unprecedented challenges on multiple fronts (see National Importance; main proposal). Drug discovery is dogged by a crippling attrition rate (>95%), and a current cost/launch of ~£1.8bn (including failed campaigns and borrowing costs). Substantial costs fall early in drug discovery because relatively few campaigns proceed to the launch of a therapeutic product. Greater innovation is urgently required to help address attrition in early-stage drug discovery.
To illustrate the value of my autonomous platform to end-users in discovery industries, I will exemplify my novel approach via the discovery of novel chemotypes for medicinally-relevant proteins. To catalyse follow-on highlight articles in high-impact journals (eg Science, Nature series, C&E News, Cell series), trade magazines/websites and drug discovery blogs, I will continue to work closely with the University's Press Office to issue press releases to coincide with high-impact publications.
The realisation of the full impact of my autonomous discovery platform will require alignment with future end-user needs. This alignment will be facilitated by partnership with GSK and AstraZeneca. In addition, I will hold a 1-day end-user workshop to which my partner, other pharmaceutical companies, SMEs, non-for-profit discovery organisations and other discovery companies will be invited: the objective will be disseminate the early results of my fellowship and to consult disparate users on their future discovery needs. Before the end of Year 4, I will visit four discovery companies to seek detailed views on fit with their specific needs, and alternative mechanisms for translation/commercialisation of my platform. In Year 5, I will commission a thorough assessment of the capabilities of competing technologies, and host three end-users to allow them to apply my platform to one of their targets. Following liaison with Leeds's commercialisation team and IP Group, I will identify the most appropriate mechanism to make my platform accessible to end-users. I will then implement a plan to ensure that my autonomous platform becomes accessible to end-users in discovery-based industries.
2. The general public
The ultimate beneficiaries of the research are the general public who benefit from the development of new medicines. Drug discovery is a topic of huge interest to the general public, and the idea of automating the process would be readily communicated to them. I will commission a cartoon aimed key stage 3 pupils that illustrates imaginatively my platform: the idea of mimicking evolution would appeal to this age group and excite them ahead of selecting GCSEs. The cartoon will be uploaded to appropriate websites (e.g. Leeds's YouTube channel) and targeted at science teachers through our ongoing schools liaison activities.
Press releases to coincide with major publications (see above) will catalyse media coverage that engages the general public. I will also continue to communicate my science through invited visits to schools, education conferences, and Café Scientifique (Headingley, Leeds).
The accompanying Pathway to Impact document provides details of how the following specific objectives will be met:
PtI1. To continue to align my fellowship with future discovery needs;
PtI2. To strengthen the perceived value of my autonomous platform for lead generation;
PtI3. To secure pathways to commercialise / translate my autonomous platform; and
PtI4. To engage the general public about my autonomous platform for lead generation.
1. Companies engaged in molecular discovery
The discovery of small molecule drugs is undertaken within the pharmaceutical sector (comprising large pharmaceutical companies) and the medical biotechnology sector (that includes smaller drug discovery companies). These companies face unprecedented challenges on multiple fronts (see National Importance; main proposal). Drug discovery is dogged by a crippling attrition rate (>95%), and a current cost/launch of ~£1.8bn (including failed campaigns and borrowing costs). Substantial costs fall early in drug discovery because relatively few campaigns proceed to the launch of a therapeutic product. Greater innovation is urgently required to help address attrition in early-stage drug discovery.
To illustrate the value of my autonomous platform to end-users in discovery industries, I will exemplify my novel approach via the discovery of novel chemotypes for medicinally-relevant proteins. To catalyse follow-on highlight articles in high-impact journals (eg Science, Nature series, C&E News, Cell series), trade magazines/websites and drug discovery blogs, I will continue to work closely with the University's Press Office to issue press releases to coincide with high-impact publications.
The realisation of the full impact of my autonomous discovery platform will require alignment with future end-user needs. This alignment will be facilitated by partnership with GSK and AstraZeneca. In addition, I will hold a 1-day end-user workshop to which my partner, other pharmaceutical companies, SMEs, non-for-profit discovery organisations and other discovery companies will be invited: the objective will be disseminate the early results of my fellowship and to consult disparate users on their future discovery needs. Before the end of Year 4, I will visit four discovery companies to seek detailed views on fit with their specific needs, and alternative mechanisms for translation/commercialisation of my platform. In Year 5, I will commission a thorough assessment of the capabilities of competing technologies, and host three end-users to allow them to apply my platform to one of their targets. Following liaison with Leeds's commercialisation team and IP Group, I will identify the most appropriate mechanism to make my platform accessible to end-users. I will then implement a plan to ensure that my autonomous platform becomes accessible to end-users in discovery-based industries.
2. The general public
The ultimate beneficiaries of the research are the general public who benefit from the development of new medicines. Drug discovery is a topic of huge interest to the general public, and the idea of automating the process would be readily communicated to them. I will commission a cartoon aimed key stage 3 pupils that illustrates imaginatively my platform: the idea of mimicking evolution would appeal to this age group and excite them ahead of selecting GCSEs. The cartoon will be uploaded to appropriate websites (e.g. Leeds's YouTube channel) and targeted at science teachers through our ongoing schools liaison activities.
Press releases to coincide with major publications (see above) will catalyse media coverage that engages the general public. I will also continue to communicate my science through invited visits to schools, education conferences, and Café Scientifique (Headingley, Leeds).
The accompanying Pathway to Impact document provides details of how the following specific objectives will be met:
PtI1. To continue to align my fellowship with future discovery needs;
PtI2. To strengthen the perceived value of my autonomous platform for lead generation;
PtI3. To secure pathways to commercialise / translate my autonomous platform; and
PtI4. To engage the general public about my autonomous platform for lead generation.
Organisations
- University of Leeds (Fellow, Lead Research Organisation)
- AstraZeneca (Collaboration)
- Max Planck Society (Collaboration)
- University of St Andrews (Collaboration)
- GlaxoSmithKline (GSK) (Collaboration)
- Rosalind Franklin Institute (Collaboration)
- AstraZeneca (United Kingdom) (Project Partner)
- Max Planck Institutes (Project Partner)
- GlaxoSmithKline (United Kingdom) (Project Partner)
People |
ORCID iD |
Adam Nelson (Principal Investigator / Fellow) |
Publications
Zhang R
(2019)
Construction of a Shape-Diverse Fragment Set: Design, Synthesis and Screen against Aurora-A Kinase.
in Chemistry (Weinheim an der Bergstrasse, Germany)
Trindade AF
(2020)
Fragment-oriented synthesis: ß-elaboration of cyclic amine fragments using enecarbamates as platform intermediates.
in Chemical communications (Cambridge, England)
Townley C
(2022)
A unified "top-down" approach for the synthesis of diverse lead-like molecular scaffolds.
in Bioorganic & medicinal chemistry letters
Schölermann B
(2022)
Identification of Dihydroorotate Dehydrogenase Inhibitors Using the Cell Painting Assay.
in Chembiochem : a European journal of chemical biology
Rice S
(2019)
Unified synthesis of diverse building blocks for application in the discovery of bioactive small molecules
in Tetrahedron
Rice S
(2021)
Efficient unified synthesis of diverse bridged polycyclic scaffolds using a complexity-generating 'stitching' annulation approach.
in Chemical communications (Cambridge, England)
Pahl A
(2023)
Morphological subprofile analysis for bioactivity annotation of small molecules.
in Cell chemical biology
Okolo E
(2023)
Scaffold Remodelling of Diazaspirotricycles Enables Synthesis of Diverse sp 3 -Rich Compounds With Distinct Phenotypic Effects
in Chemistry - A European Journal
Description | We have shown that activity-directed synthesis is a powerful approach to drive early-stage drug discovery. We have demonstrated that the approach is general, enabling many different classes of molecules to be discovered that have activity against disparate medicinally-relevant targets. We have also demonstrated that activity-directed synthesis can enable autonomous molecular discovery i.e. in which algorithms design sets of reactions and robotics execute the necessary experimental steps. |
Exploitation Route | Companies can exploit the new discovery capabilities in their drug discovery programmes. In addition, academic beneficiaries include biomedical scientists who discovery (and then exploit) chemical probes in chemical biology programmes. Leeds's partnership with the Rosalind Franklin Institute will enable the outcomes to be more widely exploited in high-throughput molecular discovery. We are aware that aspects of the activity-directed synthesis workflow are being used to drive early-stage drug discovery within academia. |
Sectors | Pharmaceuticals and Medical Biotechnology |
Description | Autonomous Discovery of Functional Small Molecules |
Amount | £1,245,808 (GBP) |
Funding ID | EP/N025652/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start | 09/2016 |
End | 09/2021 |
Description | Autonomous Phenotype-Directed Molecular Discovery |
Amount | £1,184,398 (GBP) |
Funding ID | EP/W002914/1 |
Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
Sector | Public |
Country | United Kingdom |
Start | 03/2022 |
End | 09/2025 |
Description | Phenotype-directed discovery of unnatural products |
Amount | £319,000 (GBP) |
Organisation | The Leverhulme Trust |
Sector | Charity/Non Profit |
Country | United Kingdom |
Start | 09/2018 |
End | 09/2022 |
Description | ADS |
Organisation | AstraZeneca |
Department | Research and Development AstraZeneca |
Country | United Kingdom |
Sector | Private |
PI Contribution | We have used the assets from partners in activity-directed synthesis |
Collaborator Contribution | THe partners have provided assets (protein, assay) and helped align the project with end-user need |
Impact | None to date |
Start Year | 2016 |
Description | ADS |
Organisation | GlaxoSmithKline (GSK) |
Country | Global |
Sector | Private |
PI Contribution | We have used the assets from partners in activity-directed synthesis |
Collaborator Contribution | THe partners have provided assets (protein, assay) and helped align the project with end-user need |
Impact | None to date |
Start Year | 2016 |
Description | ADS |
Organisation | Max Planck Society |
Country | Germany |
Sector | Charity/Non Profit |
PI Contribution | We have used the assets from partners in activity-directed synthesis |
Collaborator Contribution | THe partners have provided assets (protein, assay) and helped align the project with end-user need |
Impact | None to date |
Start Year | 2016 |
Description | Autonomous phenotype-directed molecular discovery |
Organisation | Max Planck Society |
Department | Max Planck Institute for Molecular Physiology |
Country | Germany |
Sector | Academic/University |
PI Contribution | This an EPSRC international centre-to-centre collaboration. Leeds is contributing diversity-oriented synthetic methods to enable phenotype-directed molecular discovery; and chemical proteomics/biology expertise. These capabilities are critical to the discovery of bioactive small molecules, and the determination of their mechanism of action. |
Collaborator Contribution | Max Planck Institute for Molecular Physiology are contributing high-throughput high-content imaging assays (cell painting assay). Rosalind Franklin Institute is contributing expertise in algorithm development and high-throughput chemical synthesis. |
Impact | To date, the collaboration has yielded new insights into the use of the cell painting assay to enable insights into biological mechanism. The collaboration is highly multidisciplinary, and involves machine learning, synthetic chemistry, cell biology and chemical biology. |
Start Year | 2022 |
Description | Autonomous phenotype-directed molecular discovery |
Organisation | Rosalind Franklin Institute |
Country | United Kingdom |
Sector | Charity/Non Profit |
PI Contribution | This an EPSRC international centre-to-centre collaboration. Leeds is contributing diversity-oriented synthetic methods to enable phenotype-directed molecular discovery; and chemical proteomics/biology expertise. These capabilities are critical to the discovery of bioactive small molecules, and the determination of their mechanism of action. |
Collaborator Contribution | Max Planck Institute for Molecular Physiology are contributing high-throughput high-content imaging assays (cell painting assay). Rosalind Franklin Institute is contributing expertise in algorithm development and high-throughput chemical synthesis. |
Impact | To date, the collaboration has yielded new insights into the use of the cell painting assay to enable insights into biological mechanism. The collaboration is highly multidisciplinary, and involves machine learning, synthetic chemistry, cell biology and chemical biology. |
Start Year | 2022 |
Description | High-throughput molecular discovery |
Organisation | Rosalind Franklin Institute |
Country | United Kingdom |
Sector | Charity/Non Profit |
PI Contribution | Synthetic methods to explore novel chemical space |
Collaborator Contribution | High-throughput chemistry capabilities |
Impact | None to date. This is a multidiscplinary partnership that combines chemistry, data science and biology |
Start Year | 2020 |
Description | MPI |
Organisation | Max Planck Society |
Department | Max Planck Institute for Molecular Physiology |
Country | Germany |
Sector | Academic/University |
PI Contribution | Screening compounds |
Collaborator Contribution | HT screening and follow-up biology |
Impact | None to date |
Start Year | 2014 |
Description | Phenotype-directed discovery |
Organisation | University of St Andrews |
Country | United Kingdom |
Sector | Academic/University |
PI Contribution | Activity-directed synthesis discovery platform |
Collaborator Contribution | Complementary expertise in parasite biology |
Impact | None to date |
Start Year | 2018 |