Dial-a-Molecule Grand Challenge Network, Phase III
Lead Research Organisation:
University of Southampton
Department Name: Sch of Chemistry
Abstract
Designed molecules are the basis of much of modern life, for example as pharmaceuticals, agrochemicals, plastics, liquid crystals, electronic materials etc). The task of discovering new molecules with desired properties is constrained by the time it takes to make them leading to compromises. For example in pharmaceuticals when the need to use simple, easy to make compounds leads to cross-activity (interaction with other than the target biological system) as the compromise, and hence undesirable side effects.
The aim of 'Dial-a-Molecule' is to make the synthesis of any desired molecule as quick as ordering an existing chemical from your supplier. A linked aim is to make production of chemicals more efficient with lower energy usage and less waste.
The Dial-a-Molecule Network was established as one of three 'Grand Challenges' for Physical Sciences by EPSRC in 2010 with the aim that "In 20-40 years, scientists will be able to deliver any desired molecule within a timeframe useful to the end-user, using safe, economically viable and sustainable processes." In its first phase it produced a roadmap (www.dial-a-molecule.com) to describe the advances needed to achieve the Grand Challenge. Strong collaborations with and input from other disciplines, such as mathematics, engineering, computer science, data science, and biology, is an essential component to realizing the aims of Dial-a-Molecule and in the second phase it progressed in building the multi-disciplinary community needed to tackle the challenge. It also started to promote a fundamental change in the way that synthetic chemistry is undertaken towards becoming a more data-centric discipline, embracing the data revolution and advances in computing power and technology that has occurred over the last 10 years.
The current grant application seeks funding to extend the network for a 3 year period, with clear plans to be self-sustaining in some form at the end. Although the overall objectives remain unchanged, plans for Phase III of the network will focus on delivering impact in in six key themes covering changing way synthesis is carried out with a particular emphasis on data capture and use, developing new ways of carrying out and studying reactions with an emphasis on automation, inventing new reactions with a potentially transformative effect on the challenge, and establishing a UK infrastructure and profile in the area. The network will continue to promote inter-disciplinary mobility, and will put a strong emphasis on supporting Early Career Researchers. The network has been particularly successful in industrial engagement; this will be further developed through initiatives such as the Synthesis College and SME networks.
The network will organize a large number of open meetings aimed at bringing together different disciplines to discuss specific topics, identify opportunities, and establish new collaborations to seek funding. It will also directly facilitate applications for funding in the area by supporting meetings to refine ideas and consortia, supporting visits between partners, and funding small proof of concept projects. It will provide training in important areas, and facilitate the introduction of new techniques (software and automation) to help equip the next generation of researchers with necessary skills. The Network has the funding of three 'headline' projects as specific aims in phase III - a 'Dial-a-Molecule' Grand Challenge Institute to provide a national focus and critical mass of people and equipment; a national collection of catalysts and ligands; and a demonstration 'Dial-a-molecule' machine.
The aim of 'Dial-a-Molecule' is to make the synthesis of any desired molecule as quick as ordering an existing chemical from your supplier. A linked aim is to make production of chemicals more efficient with lower energy usage and less waste.
The Dial-a-Molecule Network was established as one of three 'Grand Challenges' for Physical Sciences by EPSRC in 2010 with the aim that "In 20-40 years, scientists will be able to deliver any desired molecule within a timeframe useful to the end-user, using safe, economically viable and sustainable processes." In its first phase it produced a roadmap (www.dial-a-molecule.com) to describe the advances needed to achieve the Grand Challenge. Strong collaborations with and input from other disciplines, such as mathematics, engineering, computer science, data science, and biology, is an essential component to realizing the aims of Dial-a-Molecule and in the second phase it progressed in building the multi-disciplinary community needed to tackle the challenge. It also started to promote a fundamental change in the way that synthetic chemistry is undertaken towards becoming a more data-centric discipline, embracing the data revolution and advances in computing power and technology that has occurred over the last 10 years.
The current grant application seeks funding to extend the network for a 3 year period, with clear plans to be self-sustaining in some form at the end. Although the overall objectives remain unchanged, plans for Phase III of the network will focus on delivering impact in in six key themes covering changing way synthesis is carried out with a particular emphasis on data capture and use, developing new ways of carrying out and studying reactions with an emphasis on automation, inventing new reactions with a potentially transformative effect on the challenge, and establishing a UK infrastructure and profile in the area. The network will continue to promote inter-disciplinary mobility, and will put a strong emphasis on supporting Early Career Researchers. The network has been particularly successful in industrial engagement; this will be further developed through initiatives such as the Synthesis College and SME networks.
The network will organize a large number of open meetings aimed at bringing together different disciplines to discuss specific topics, identify opportunities, and establish new collaborations to seek funding. It will also directly facilitate applications for funding in the area by supporting meetings to refine ideas and consortia, supporting visits between partners, and funding small proof of concept projects. It will provide training in important areas, and facilitate the introduction of new techniques (software and automation) to help equip the next generation of researchers with necessary skills. The Network has the funding of three 'headline' projects as specific aims in phase III - a 'Dial-a-Molecule' Grand Challenge Institute to provide a national focus and critical mass of people and equipment; a national collection of catalysts and ligands; and a demonstration 'Dial-a-molecule' machine.
Planned Impact
The Dial-a-Molecule Grand Challenge network builds capability though engaging people across a wide range of disciplines.
The network directly benefits members (industrial and academic) in many ways: the opportunity to publicise their work or products; learn about other work, particularly from other disciplines; to develop collaborations, again involving other disciplines or between academia and industry leading to new funding opportunities; provides the impetus to develop new research areas or products, again potentially attracting funding; as well as potential access to small sums of money for conference attendance, visiting other laboratories, or proof of concept studies.
It benefits Industry by guiding academic research towards problems of current interest, and by making it more aware of current developments in areas of importance.
It benefits people by extending their skill set through training in new methods so improving opportunities for jobs and personal development. Examples include use of electronic data capture, and the use of statistical, physical organic and chemical engineering tools.
The scientific and technical advances the network aim to promote in the production of molecules 'to order' while minimizing the environmental footprint associated with the synthesis have immense potential impact:
Knowledge Impact: Improvements in chemical synthesis through scientific advances in the reactions developed, equipment used, and the way data is obtained and used will allow anyone in academia or industry involved in chemical synthesis, or the applications of chemical entities to progress more quickly.
Economic Impact: The development of new techniques, strategies and ways of working will accelerate the development of new/improved products for end users improving competitiveness. It should also lower the economic and environmental cost of chemical production.
The enhanced skill set of synthetic chemists Dial-a-Molecule promotes, and the establishment of a world leading Grand Challenge Institute, will both provide incentive for companies using advanced chemical synthesis to locate in the U.K.
Dial-a-Molecule will drive the development of new products (equipment, software and consumables) from the supply chain allowing the establishment and growth of companies producing them. If the model for new chemical entity delivery proposed by the 'synthesis machine' theme proves economically viable, it provides a new chemical synthesis business model.
Societal Impact:
Policy. Dial-a-Molecule aims to influencing policy on the future of chemical discovery and manufacture and argues for the importance of increased government investment in UK science by demonstrating the societal and economic benefits that will accrue from its programs.
Health. Advances in next-generation healthcare (personalized medicine, diagnostics, theragnostics, targeted therapies, drug delivery, modified biologies etc) will be driven by chemical technologies arising from Dial-a-Molecule. As drugs are designed to interact with only the receptor of interest, or are tuned to the specific genetic make-up of an individual, the requirement for rapid synthesis of complex molecules will spiral to the point where the Dial-a-Molecule technologies are essential for progress.
Environment: Many facets of Dial-a-Molecule have as their goal more sustainable processes/products with reduced environmental footprint and energy usage. One example is to developing catalysts using more readily available and less toxic metals than those currently used. Another is the drive towards data-driven study and optimization of reactions to give more efficient processes.
Overall we believe that the work promoted by the network will have considerable scientific, commercial and societal impact. An important function of the network is to provide a pathway to impact for such work and we believe we have the mechanisms in place to enable it.
The network directly benefits members (industrial and academic) in many ways: the opportunity to publicise their work or products; learn about other work, particularly from other disciplines; to develop collaborations, again involving other disciplines or between academia and industry leading to new funding opportunities; provides the impetus to develop new research areas or products, again potentially attracting funding; as well as potential access to small sums of money for conference attendance, visiting other laboratories, or proof of concept studies.
It benefits Industry by guiding academic research towards problems of current interest, and by making it more aware of current developments in areas of importance.
It benefits people by extending their skill set through training in new methods so improving opportunities for jobs and personal development. Examples include use of electronic data capture, and the use of statistical, physical organic and chemical engineering tools.
The scientific and technical advances the network aim to promote in the production of molecules 'to order' while minimizing the environmental footprint associated with the synthesis have immense potential impact:
Knowledge Impact: Improvements in chemical synthesis through scientific advances in the reactions developed, equipment used, and the way data is obtained and used will allow anyone in academia or industry involved in chemical synthesis, or the applications of chemical entities to progress more quickly.
Economic Impact: The development of new techniques, strategies and ways of working will accelerate the development of new/improved products for end users improving competitiveness. It should also lower the economic and environmental cost of chemical production.
The enhanced skill set of synthetic chemists Dial-a-Molecule promotes, and the establishment of a world leading Grand Challenge Institute, will both provide incentive for companies using advanced chemical synthesis to locate in the U.K.
Dial-a-Molecule will drive the development of new products (equipment, software and consumables) from the supply chain allowing the establishment and growth of companies producing them. If the model for new chemical entity delivery proposed by the 'synthesis machine' theme proves economically viable, it provides a new chemical synthesis business model.
Societal Impact:
Policy. Dial-a-Molecule aims to influencing policy on the future of chemical discovery and manufacture and argues for the importance of increased government investment in UK science by demonstrating the societal and economic benefits that will accrue from its programs.
Health. Advances in next-generation healthcare (personalized medicine, diagnostics, theragnostics, targeted therapies, drug delivery, modified biologies etc) will be driven by chemical technologies arising from Dial-a-Molecule. As drugs are designed to interact with only the receptor of interest, or are tuned to the specific genetic make-up of an individual, the requirement for rapid synthesis of complex molecules will spiral to the point where the Dial-a-Molecule technologies are essential for progress.
Environment: Many facets of Dial-a-Molecule have as their goal more sustainable processes/products with reduced environmental footprint and energy usage. One example is to developing catalysts using more readily available and less toxic metals than those currently used. Another is the drive towards data-driven study and optimization of reactions to give more efficient processes.
Overall we believe that the work promoted by the network will have considerable scientific, commercial and societal impact. An important function of the network is to provide a pathway to impact for such work and we believe we have the mechanisms in place to enable it.
Organisations
Publications
Bourne R
(2019)
Introduction to Synthesis 4.0: towards an internet of chemistry
in Reaction Chemistry & Engineering
Cherkasov N
(2020)
Monitoring Chemistry In Situ with a Smart Stirrer: A Magnetic Stirrer Bar with an Integrated Process Monitoring System.
in ACS sensors
Donald JR
(2019)
Radical cyanomethylation via vinyl azide cascade-fragmentation.
in Chemical science
Gale EM
(2020)
Improving reaction prediction.
in Nature chemistry
Gale EM
(2021)
Author Correction: Improving reaction prediction.
in Nature chemistry
Jones CD
(2021)
High-Yielding Flow Synthesis of a Macrocyclic Molecular Hinge.
in Journal of the American Chemical Society
Magné V
(2019)
Synthesis of Air-stable, Odorless Thiophenol Surrogates via Ni-Catalyzed C-S Cross-Coupling.
in Chemistry (Weinheim an der Bergstrasse, Germany)
Martínez-Carrión A
(2019)
Kinetic Treatments for Catalyst Activation and Deactivation Processes based on Variable Time Normalization Analysis
in Angewandte Chemie
Martínez-Carrión A
(2019)
Kinetic Treatments for Catalyst Activation and Deactivation Processes based on Variable Time Normalization Analysis.
in Angewandte Chemie (International ed. in English)
Ollerton K
(2021)
Enabling Technology for Supramolecular Chemistry.
in Frontiers in chemistry
| Description | As a network research has not been funded directly so does not lead directly new discoveries / findings. The Network has strongly influenced the development of new collaborations and grant proposals through the many meetings it has run. Perhaps the key outcome over this period has been the huge increase in interest in one of main themes promoted by the network - the use of automation and AI to drive synthesis, particularly demonstrated by the large number of international conferences in the area over the past couple of years compared with very few - mostly run by us - before, and several large grants awarded. |
| Exploitation Route | The outcomes (which are mostly new collaborations and research directions developed by participants in our events, or readers of our Web-site and roadmap) are being developed in a number of grants, almost all in collaboration with industry. |
| Sectors | Agriculture, Food and Drink,Chemicals,Environment,Manufacturing, including Industrial Biotechology,Pharmaceuticals and Medical Biotechnology |
| URL | http://www.dial-a-molecule.org |
| Description | During the 3.5 years of the Network grant we have held 16 very well attended meetings which have lead to many new collaborations, both between academics and between academia and industry. At least 11 new EPSRC awards have been funded that owed something to the network. With strong network support an EPSRC grant to fund the establishment of the Rapid On-Line Analysis of Reactions (ROAR) centre at Imperial College was awarded and this centre is now open and recently received phase 2 funding. The network has also been presented to two international meetings, and the PI has given interviews on the topic to several magazines demonstrating the strong current interest in automated synthesis and the use of data and machine learning to guide synthesis, and the influence the network is having on the discussion. The Co-I edited a collection of papers in 'Reaction Chemistry and Engineering' on 'Synthesis 4.0' inspired by the Network aims. Several other publications have resulted from 'proof of concept' grants made by the Network. It is notable that several areas first promoted by the network, and where the network held the first meetings (particulalry several aspects of AI in Chemistry) are now substantial areas of research with much directed funding, as well as the topic for numerous meetings each year. |
| First Year Of Impact | 2019 |
| Sector | Agriculture, Food and Drink,Chemicals,Manufacturing, including Industrial Biotechology,Pharmaceuticals and Medical Biotechnology |
| Impact Types | Economic |
| Description | A Fully-Automated Robotic System For Intelligent Chemical Reaction Screening |
| Amount | £201,478 (GBP) |
| Funding ID | EP/S009965/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 01/2019 |
| End | 12/2020 |
| Description | A nitrenoid strategy to access sp3-rich nitrogen heterocycles |
| Amount | £444,750 (GBP) |
| Funding ID | EP/V061690/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 07/2021 |
| End | 09/2024 |
| Description | Artificial and Augmented Intelligence for Automated Scientific Discovery |
| Amount | £1,014,318 (GBP) |
| Funding ID | EP/S000356/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 07/2018 |
| End | 06/2021 |
| Description | Centre for Rapid Online Analysis of Reactions (ROAR), Phase II |
| Amount | £231,295 (GBP) |
| Funding ID | EP/V029037/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 02/2021 |
| End | 01/2023 |
| Description | Chemobots: Digital-Chemical-Robotics to Convert Code to Molecules and Complex Systems |
| Amount | £5,034,016 (GBP) |
| Funding ID | EP/S019472/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 02/2019 |
| End | 01/2024 |
| Description | Cognitive Chemical Manufacturing |
| Amount | £2,007,486 (GBP) |
| Funding ID | EP/R032807/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 09/2018 |
| End | 09/2022 |
| Description | EPSRC Responsive Mode. Developing Continuous Electro-Organic Catalysis - It's Got Potential (LC Morrill, T. Wirth, DL Browne) |
| Amount | £396,000 (GBP) |
| Funding ID | EP/R006504/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 11/2017 |
| End | 11/2020 |
| Description | EPSRC responsive mode. Multifunctional Electrochemical Flow Platform for High-Throughput Synthesis & Optimisation of Catalysts (C Willans, BN Nguyen, RA Bourne) |
| Amount | £741,347 (GBP) |
| Funding ID | EP/R009406/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 01/2018 |
| End | 03/2022 |
| Description | Engineering and Physical Sciences Research Council Network Dial-a-Molecule Phase III |
| Amount | £254,038 (GBP) |
| Funding ID | EP/P007589/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 10/2016 |
| End | 09/2019 |
| Description | FLEXICHEM: Flexible Digital Chemical Manufacturing Through Structure/Reactivity Relationships |
| Amount | £1,201,544 (GBP) |
| Funding ID | EP/V050990/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 01/2022 |
| End | 12/2024 |
| Description | Reactors and Reproducibility: Advancing Electrochemistry for Organic Synthesis |
| Amount | £372,708 (GBP) |
| Funding ID | EP/T001631/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 10/2019 |
| End | 09/2022 |
| Description | Responsive Mode. Combining Chemical Robotics and Statistical Methods to Discover Complex Functional Products (A Lapkin, D Woods, L. Cronin) |
| Amount | £1,227,510 (GBP) |
| Funding ID | EP/R009902/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 02/2018 |
| End | 01/2021 |
| Description | Royal Society University Research Fellowship: Control of non-covalent synthesis and dynamic equilibria in flow |
| Amount | £593,928 (GBP) |
| Organisation | The Royal Society |
| Sector | Charity/Non Profit |
| Country | United Kingdom |
| Start | 01/2021 |
| End | 12/2026 |
| Description | Standard Research (Strategic equipment). Rapid Online Analysis of Reactions (ROAR) |
| Amount | £2,757,688 (GBP) |
| Funding ID | EP/R008825/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 01/2018 |
| End | 12/2020 |
| Description | Water as synthetic reaction medium: realising its green chemistry credential |
| Amount | £871,781 (GBP) |
| Funding ID | EP/S013768/1 |
| Organisation | Engineering and Physical Sciences Research Council (EPSRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 03/2019 |
| End | 08/2022 |
| Description | AI for Reaction Outcome and Synthetic Route Prediction |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | International conference / workshop organised by Dial-a-Molecule in collaboration with the Directed Assembly and AI3SD networks on the topic of AI for Reaction Outcome and Synthetic Route Prediction. It was held over 3 days (March 9th -11th 2020) at the DeVere Totworth Court Hotel, Gloucestershire. There were 30 speakers from both Industry and academia. Several of the presentations had to be given on-line due to COVID inspired travel restrictions (though the conference finished before any were imposed in the U.K. The meeting was supported by around 18 exhibitors. As well as the official conference write-up (doi:10.5258/SOTON/P0021) it was the subject of a write-up in Nature Chemistry (Nat. Chem. 12, 509-510 (2020).) |
| Year(s) Of Engagement Activity | 2020 |
| URL | https://eprints.soton.ac.uk/441628/ |
| Description | AI/Machine Learning for Chemical Discovery and Development |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | The discussion was directed at targeted areas and key challenges for wider application of AI/ML to chemical research. We started with reflection on what chemists are good at, and what chemists are not. This was based on the consensus view that AI/ML should be used to assist chemists, not to replace them. We concluded that chemists are good at: Normal retrosynthetic analysis (but habitual use of certain reactions is a problem) Recognising large trends Predicting standard reactivities Intuitive extrapolation The attendees agreed that chemists/human are poor at: Predicting complex effects of small changes or properties of mixtures and blends Handling large volume of data Quantifying observations and predictions Covering chemical and operational space systematically Publishing data in useful formats Including sensors/detectors for data Based on this discussion, a number of potential research areas for discussion in Day 2 were suggested by the attendees: Data preparation (experimental errors) - dealing with bias - access data - standard Smaller datasets Confidence evaluation/benchmarking/generating chemical understanding Representation of molecules and reactions/interactions Route design with constraints/ranking/reaction outcomes Molecular discovery with multiple objectives Reaction/reagent/catalyst optimisation/impurities Predicting crystallisation/isolation Engagement between disciplines These were followed by interest identification of research areas over lunch period and voting later on the day. Discussion on AI/ML methodology: The discussion was based on three areas. Method/problem suitability/benchmarking standards: The difficulty in providing benchmarks across the field was highlighted. However, a small number of high quality benchmarking datasets, which are common in the AI/ML community and may be hosted by AI3SD, will be well received in this area of research. These will lower the barrier for researchers from other disciplines, g.mathematics and computer science, to joint our field. A common issues in comparing studies with similar objectives is disparity in chemical space coverage of the training sets, which prevents fair comparison. Thus, measurement of diversity in the datasetmust be included. The approach and consideration toward preparing benchmarking datasets will set the standard for researchers and reviewers in this field. Going forward, community-built data standard, e.g.Stepstone project, should be widely adopted as a requirement for grant funding and publication in chemistry journals. This will need leadership and vision from UKRC and the RSC. Benchmarking and metrics toolsto compare models are also required in both academia and industry. An example of these is GuacaMol, an open source python package developed by BenevolentAI for benchmarking of models for de novomolecular design. While it is appreciated that each research question will require adjustment to the AI/ML methodology, the attendees suggested that some guidelines/standard practicewill be helpful for newcomers to the field. These include weighing methods for noisy data, recommended on descriptors : datapoints ratio, balancing between exploration and cost, and standard pairings between problems and AI/ML methods. Dealing with non-perfect data, chemical bias, and causal relationship: The attendees generally agree that in chemical context, the current limitation is the data rather than the AI/ML methodology. Due to bias in publication system, the available data in the literature can be incomplete (focused on yield, which consists of several types, instead of reactivity and kinetic data) and biased (focused on successes and not negative results). Data collection by chemists has been non-systematic, sometimes without context (g.reactor type, temperature, reaction time, purity of reagents and solvent, etc.) which leads to the need for curation. Certain progresses have been made in these areas, e.g.Inchi code and Stepstone project. However, there is a clear need for leadership in setting the standards and formats for data reporting(manuscript and ESI), including standard language, metadata and ontology, which will lead to usable data for AI/ML and open knowledge exchange. The adopted format also needs flexibility for expansion as we do not know what data we need to capture in the future. Motivation for data deposition, beyond the goodwill of the research, needs to be provided. Despite all these drawbacks, there is a large volume of useful data, e.g.physical properties and reactivity, in the literature and tools to make some machine readable. Reaction data can also be useful as starting points if due consideration is given to its inherent bias and incompleteness. These are a potentially valuable sources of data, which do not have complicated legal barrier to assess. Further investment in data extraction tools and data curationis important for the field to move forward. Some funders do fund community-owned-database development, but not the EPSRC. Perhaps exception can be made given its importance. Confidence and uncertainty: This topic generated a great deal of interest in the room. The consensus is that confidence and uncertaintyare critical and must be put front and centre in the use of AI/ML in chemical research. This will allow informed decision with expert interpretation of AI/ML output. In the context of synthetic route design, they can be used to rank predictions and establish safety margin. Linking uncertainty to variation in input will also provide critical understanding of the process. This is particularly important when evaluating outliers. The required accuracy of the model, however, is application dependent and maximum accuracy is not always the objective. It is important for researchers to recognise that the maximum accuracy of a model should not exceed the amount of noise in training data. There is a cultural issue with reporting errors in experimental data in synthetic science. Control studies, including DoE, may be employed to evaluate the confidence in the model when required. Background knowledge may also be included in the model to address uncertainty and smaller datasets. Future model may see merging between mechanistic models and AI/ML models towardadaptive modelswhich can improve as strategic data become available. Discussion on data What is 'good' data in your area of interest? There is a diverse range of need depending on the application and given that any data is better than none, no specific barrier was proposed. Generally metadataincluding context and experimental details is required. What to do about old data? This was covered above. Can industrial data be shared? Yes, under certain circumstances. Lhasa has been acting as a trusted intermediary to help company sharing information and models based on in-house data. Companies can share info on patented molecules. These may be further facilitated with encryption and standardised data structures and sharing agreements. It is important for Industry to see solid value proposition and to recognise value in the data they hold through the right proof-of-concept projects. Physical properties data are probably easier to share compared to reaction data. Topics identification through voting: After the discussion on AI/ML methodology and data the topics proposed during lunch break (using 3 post-it's per academics) were put to an informed vote. Those which have more than 1 post-it's were voted on and the top three were taken forward as discussion topics for Day 2. The results are summarised below. Model interpretability 11 Route design/ranking with constraints 9 Data mining/access/standard 7 Reaction/process outcome prediction and optimisation 15 Predicting reactivity and properties 13 Molecular discovery 9 Representation of molecules and interaction 10 Bias in data and capturing chemical intuition 5 On Day 2, the workshop was divided into three groups to discuss each of the selected topics in three 1-hour sessions. Attendees were able to freely move between topics as they wish. The guidance was to focus on the challenges and milestones in these areas of research. Discussion of milestones for research area 1: Model interpretability Three prompts were used to guide the discussion: Accessible tools/protocols to gain understanding Relevant descriptors/representation of molecules/reaction/ interaction Evaluation of confidence The attendees highlighted the need to understand how the AI/ML model works in order to get the best prediction and balance any inherent bias in the data. AI/ML methods in research and publications should be subjected to the same evaluation as human led studies. The interpretability of the model is essential in gaining confidence in the model, refining it, knowing its limits and understanding the relationship between physical properties of the system. To ensure these, the training data should be relevant and randomised. AI/ML models are built for interpolation and extrapolation must be approached with due consideration. Automation and high throughput experimentationshould be used to assist model interpretation, refinement and validation. Furthermore, descriptive models are sometimes better for interpretability and bias analysis than predictive models. The use of standard datasetsfor comparison between methods is highly valued. In this context of gaining confidence, evaluation of AI/ML against real world problems, instead of idealised ones, is important. Propagation of errors needs to be considered in this context. It is important to understand the research area, its rules and biases, and to interpret the research question mathematicallyfor AI/ML. This includes multi-objective optimisation, which is common in many chemical research areas. Discussion of milestones for research area 2: Reaction/process outcome prediction and optimisation Five prompts were used to guide the discussion: Predicting outcomes (product, impurities, selectivities) Optimising with gradient and without gradient AI/ML methodologies for different data types/sizes Cumulative/adaptive models which grow with data? How to approach reactivity and selectivity Questions were raised on the benefit of AI/ML over ab initio/DFT methods in this context, particularly when one needs to make predictions outside the chemical spacecovered by the training data. A suitable method is suggestions by AI/ML, which can be evaluated and decided on by chemists. Again, different level of accuracy in prediction or different questions can be used in different context, e.g.is the reaction going to be clean or messy? A more physical approach, merging mechanistic models and AI/ML models, is to predict relative reactivityof different functional groups in starting materials, which can be used to predict reaction success as a meter/index. Computational reaction network is another approach using AI/ML to predict organic reaction outcomes. An added benefit of employing AI/ML in this field is that green chemistry constraints can be built into the system. Development of relevant descriptors for these is still a challenge in this field and standard approaches to descriptors development, including stereochemistry, will be useful for new researchers. Reaction conditions and operational space are often limited by non-chemical factors. Training data do not necessarily reflect reactivity and future collection of kinetic/reactivity data, e.g.HPLC trace at t1/2with internal standard, in some form, in standardised formats, is critical. Lack of negative resultsin the literature is also an issue, which should be addressed going forward and through mining open-access theses. A holistic and multi-objective approach to include workup, purification and crystallisation may be of more importance in production context. Thismay include sustainability, yield, cost, robustness, by-product formation and optimisation of a global route instead of individual steps in a synthesis. Transferability of process between reactors/sites is a common problem. Mixtures of solvents. Given the important of this area of research, community challenges and competitionsin combination with benchmarking datasetsmay significantly speed up its development. Discussion of milestones for research area 3: Predicting reactivity and properties Five prompts were used to guide the discussion: What properties are important and challenging? How to represent biological activity? Will it work in formulation? Accuracy and cost vsfirst principles techniques Reliable benchmarks Solubilitydominates the discussion on chemical properties. This includes ionic compounds, salts, non-aqueous solvents, solvent mixtures. These are linked with formulation, co-crystals an aggregation issues in purification. Formulation, in particular, still heavily rely on experience at the moment. Descriptor development in this area must engage with experts to capture their approach through a theory-guided selection. Another property of interest to chemical development is stability of compounds, e.g. APIs, against light, oxygen, water and themselves. Underpinning many of these predictions, including solid phase/material properties, is crystal structure and lattice energy, which is still currently not accurately predicted. Properties of ionic liquid is another area where AI/ML may help. Predicting biological activityis even more complicated, due to its multi-facet and multi-stop nature, e.g. oral availability, transport across membranes, metabolism, protein binding, clearance rate, etc. Minor conformers must be considered. Mathematical description of the binding process, e.g.binding constant, structural and functional changes, is non-trivial. However, data availability and quality is less problematic in this area. The attendees suggested the use of AI/ML to develop standardised fingerprint tests, improve hit rate, and generating potential APIs from fragment data will be important. Part of these are being carried out by Benevolent AI. Some biological targets, however, are more difficult due to lack of structural data. Controlling protein-protein interactionis still currently very difficult dues to the number of possibilities, but is perhaps an important area for AI/ML in the future. Most of these targets have been identified for decades. First principles techniques have not been able to effectively address them, but AI/ML has provided success in certain cases such as predicting genotoxicity. Discussion on wider engagement Three prompts were used to start this discussion. Lhasa presented a short summary of their experience working in this field before wider discussion between all attendees. Missing expertise Training pipeline Wider perspectives on application of AI/ML to chemical research and development The workshop recognised the lack of female attendee. Wider engagement with AI/ML specialists outside chemistry has also been difficult. There is no clear undergraduate training pipeline for future scientists with sufficient training in both chemistry and data science. A number of activities were proposed to address these: Summer school (perhaps organised by AI3SD) to provide training to those who wish to be involved in this field and promote standard tools/benchmarks. Secondments for data scientists and chemists Community hub for data and code and additional training in the style of 'software carpentry' and 'data carpentry' websites/communities. Showcase data/model and competitions to improve trust in the methodology from more traditional chemists. |
| Year(s) Of Engagement Activity | 2019 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/wp-content/blogs.dir/sites/50/2019/07/AI_ML_out... |
| Description | Dal-a-Molecule Synthetic Chemistry Academic - Industrial Ideas Exchange |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | The event comprised discussion groups and workshops to identify research topics that are of mutual interest to industrial and academic partners and that could be developed into tangible projects. We aimed to develop funding strategies and identify project partners to develop viable proposals that are capable of attracting significant levels of research funding. Aims of the Event: To promote greater academic engagement with industrially-relevant pre-competitive research; To leverage support and funding for grant applications; Promote dialogue between governmental and industrial funders; Team-building - Connecting academic and industrial partners in areas of research where there is mutual interest and complimentary expertise. |
| Year(s) Of Engagement Activity | 2017 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/academic_industrial_ideas_exchange/ |
| Description | Dial-a-Molecule (Professor Richard Whitby) presentation at a meeting on "Trends in Chemical Production" organised by the Organisation for the Prohibition of Chemical Weapons, Zagreb 4-5th Oct 2017. |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Policymakers/politicians |
| Results and Impact | The talk described the EPSRC funded 'Dial-a-Molecule' Grand Challenge (www.dial-a-molecule.org) network which has the vision: "In 20-40 years, scientists will be able to deliver any desired molecule within a timeframe useful to the end-user, using safe, economically viable and sustainable processes." The Network was established through a competitive process with strong support from industry, and has around 50% non-academic membership. The network has run many meetings over the past 6 years, mostly hosted at industrial sites, aimed at bringing together the disparate disciplines needed to tackle the Grand Challenge. A roadmap has been published which identifies many of the advances needed, of which the most fundamental is that organic synthesis needs to change to a data driven discipline with much great use of automation and computation. |
| Year(s) Of Engagement Activity | 2017 |
| URL | https://www.opcw.org/news/article/opcw-scientific-advisory-board-reviews-technological-developments-... |
| Description | Dial-a-Molecule Annual Meeting 2017, Liverpool: Transforming Chemical Synthesis |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | Attendees were treated to a great set of talks, looking at how we can push the boundaries of chemical synthesis and also took advantage of the networking opportunities provided by the event. Presentations from Lee Cronin, Steve Hilton, Varinder Aggarwal and Andy Cooper covered the connected laboratory, intelligent autonomous systems, 3D printing, molecular assembly lines and computationally led design of functional molecules, demonstrating just how much is possible by forward looking and creative groups. Coleen Robinson highlighted the importance of engaging the end-user in the design and selection of automated equipment options, in order to enable uptake of these tools, whilst Mimi Hii introduced the concept of a Dial-a-Molecule Grand Challenge Institute providing widely accessible, state-of-the art equipment and expertise to accelerate the progress and capability of UK synthetic chemistry. In the final session Patrick Courtney and Bill Whittingham gave us insights into Future Opportunities, both from the European perspective and in terms of priorities for industry. The day was completed with flash presentations from ECRs and an exhibition. |
| Year(s) Of Engagement Activity | 2017 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/liverpool_ann_meet/ |
| Description | Dial-a-Molecule Annual Meeting 2018 |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | The Dial-a-Molecule Annual Meeting 2018, Enabling Synthesis, took place on 9th-10th July 2018 in the just completed iHUB, at Imperial College's new White City Campus. The main theme was the use of technology - hardware and software - to enable synthesis It included an introduction to the new Centre For Rapid Online Analysis Of Reactions (ROAR): The Dial-a-Molecule Institute@Imperial. Keynote talks such as 'Kinetic Profiling of Catalytic Organic Reactions as a Mechanistic Tool' (Prof Donna Blackmond, Scripps Research Institute), 'Multi-phasic Continuous Flow Reactors and Reactions' (Prof. John Blacker, iPRD, University of Leeds), 'Enabling Data Rich Experimentation and associated Data Analytics' (Dr. Henry Dubina, Mettler-Toledo), 'Acceleration of the Drug Discovery Process by Enabling Chemistry Technology' (Dr. Steve Djuric, Abbvie), 'Programmable Organic Synthesis for the Age of Automation' (Prof. Jeff Bode, ETH Zurich), were supported by a wide range of shorter talks. There was also a substantial exhibition, and a panel discussion. |
| Year(s) Of Engagement Activity | 2018 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/annual-meeting-2018-agenda/ |
| Description | Dial-a-Molecule Annual Meeting 2019 |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | The Dial-a-Molecule Annual Meeting 2019 was held on3-4thJuly, 2019 at the University of York The Annual Meeting of the Dial-a-Molecule network for 2019 was an opportunity to celebrate the achievements of Dial-a-Molecule and to look forward to exciting future developments. This 2-day meeting covered many areas within the Dial-a-Molecule remit as invited talks, oral presentations and posters. Breakout sessions provided the opportunity to shape the future of the network, whilst a wide ranging exhibition and ample time for networking completed the programme. |
| Year(s) Of Engagement Activity | 2019 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/annmeet2019/ |
| Description | Dial-a-Molecule Enabling Technologies Summer School 2018 |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Postgraduate students |
| Results and Impact | 24 PhD students from 18 different institutions gathered at Loughborough University's STEMLab facility 30th July - 3rd August 2018 for the first Dial-a-Molecule Summer School. The summer school began with an introduction to statistical Design of Experiments (DoE) then learning about flow chemistry, before spending time in the lab putting the DoE and the flow chemistry together, with 6 groups each running an 18 experiment design.Although this experiment used readily available kit (tubing and syringe pumps, with offline sampling) during the week the participants were also shown more sophisticated automated equipment together with options for on-line monitoring using IR, NMR or UV. The importance of both DoE and real-time analytics was reinforced using case studies from industry illustrating the depth of process understanding and control that can be gained. An overview of the different methods of 3D printing (or, more accurately, additive manufacturing) was followed by a session in the computer room introducing CAD software, allowing the students to come up with their own designs for printing. The final topic of the Summer School covered novel reactor technologies. This included the use of interactive apps to show how different permutations of flow reactors affect residence time. An overview of the different types and fabrication of laboratory reactors was used to demonstrate how reactor design can be tailored to chemical transformations. |
| Year(s) Of Engagement Activity | 2018 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/summer_school_18/ |
| Description | Dial-a-Molecule Summer School 2019 |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Postgraduate students |
| Results and Impact | Dial-a-Molecule Summer School 2019 29th July - 2nd August, 2019 Loughborough University The second Dial-a-Molecule Summer School in Enabling Technologies was held during the week of 29th July at Loughborough University. This year, as well as PhD students from the UK, the event attracted 2 students from overseas and 3 young researchers from industry. An initial short introduction from each participant revealed a diverse range of backgrounds and research interests, setting the scene for an interesting week. The summer school began with an introduction to statistical Design of Experiments (DoE), with the first evening being spent putting the lessons into practice by designing and making paper helicopters, whilst enjoying a drink. Next, it was on to learning about flow chemistry, before spending time in the lab putting the DoE and the flow chemistry together. The 18 experiment design was carried out in an afternoon, using readily available kit (tubing and syringe pumps) which the 6 teams assembled themselves. HPLC samples were collected for off-line analysis and were run overnight. After reviewing the data, DoE software was used (together with chemical knowledge) in a hands-on session to analyse, visualise and draw conclusions from the results. Whilst waiting for the LC results, the importance - and the scope - of DoE was illustrated with some powerful examples from industry. This showed the depth of process understanding and control that can be gained. The range of analytical tools available for on-line, at-line or in-line measurements was also discussed and their use within automated chemistry and closed loop optimisation was presented. Participants had the opportunity to see some of the monitoring equipment in the lab, including in-line IR making use of a 3D printed adapter. An overview of the different methods of 3D printing (or, more accurately, additive manufacturing) was followed by a session in the computer room introducing CAD software, taking the students through the steps to design their own 'conical flask keyring'. 3D Printers were then seen in action, making the keyrings as an exclusive memento for everyone to take away!Software currently under development to allow very straightforward design of flow chemistry chips was then presented by Added Scientific and a 'flow reactor in a suitcase', made to a large extent from 3D printed parts and operated by compressed air, was demonstrated by Matt Penny of UCL. The final topic of the Summer School covered novel reactor technologies, particularly in relation to flow chemistry. The utility of CSTRs was discussed as an alternative to continuous chemistry in tubes and use of the fReactor lab scale cascade CSTR was exemplified for multi-phasic reactions. |
| Year(s) Of Engagement Activity | 2019 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/summerschool19/ |
| Description | Dial-a-Molecule and Directed Assembly Networks ECR Event: Supporting Synthesis and Self-Assembly |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | This event brought together the Early Career Researchers from the Dial a Molecule (DaM) and Directed Assembly (DA) Networks, with a focus on how enabling technology can accelerate the aims of both Networks. It included flash presentations, academic talks, industrial presentations, career development sessions, and time to discuss collaborations and future grant proposals, with a strong emphasis on cross-network activity. Topics covered included: automation; organic synthesis; self-assembly; functional materials; flow chemistry; high-throughput methodology, and integrating new technologies into the lab. |
| Year(s) Of Engagement Activity | 2017 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/ecr_liverpool/ |
| Description | Dial-a-Molecule, data, and academic adoption of ELNs. Talk by Prof. Richard Whitby at the Beilstein Institute 'Open Science and the Chemistry Lab of the Future' meeting 22nd-24th May, 2017, Frankfurt. |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | The EPSRC funded 'Dial-a-Molecule' Grand Challenge network (www.dial-a-molecule.org) has the vision: "In 20-40 years, scientists will be able to deliver any desired molecule within a timeframe useful to the end-user, using safe, economically viable and sustainable processes." The capability to select a synthetic route to a target, which will work reliably, is key. Nearly 50 years of attempts at Computer Aided Synthesis Design programs have shown how difficult this is. The main problem is our current inability to reliably predict the outcome of unknown reactions. Key problems are the discrete nature of molecular structures, the huge number of variables involved in performing reactions, and the very limited data which is provided through conventional publication routes. In particular typically only the most successful reactions and conditions are published - data on reactions / conditions which do not work well is largely lost, but crucial for predictions. The talk described the work Dial-a-Molecule has carried out towards the enabling steps of ensuring the data is captured in computer readable form (academic adoption of Electronic Laboratory Notebooks (ELNs)) and establishing standards to allow its effective sharing including automated extraction of data. The meeting brought together many people interested in the importance of making research data readily available and usable and will help initiate new collaborations. |
| Year(s) Of Engagement Activity | 2017 |
| URL | http://www.beilstein-institut.de/en/symposia/archive/open-science/open-science-2017 |
| Description | Dial-a-Molecule. From Spectral Data To Chemical Knowledge (Bath, 7th Nov 2017) |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | This cross disciplinary meeting focussed on the spectroscopic analysis of chemical processes, application of this data to advance chemical knowledge and the role of computers and automation. The meeting included topics such as how to handle and manipulate multiple data streams acquired concurrently, direct comparison of reaction profiles form different techniques, archiving and data security, the use of self-optimising reactors and scientific data management. Many vendors also exhibited during the meeting, to enable users and vendors to share their thoughts on current issues and solutions alike. Vendors include ACD/Labs, Mestrelabs, Mettler-Toledo, TGK, Bruker, S-PACT and Advion. The meeting was jointly organised by the RSC Molecular Spectroscopy Group, the RSC Chemical Information and Computer Applications Group (CICAG) and the Dial-a-Molecule EPSRC Grand Challenge Network. Talks included: Julian Morris (CPACT) "Advances and Challenges in Process Analytical Technologies and Predictive Analytics in Process Development and Manufacturing" Yui Tateno (GSK) "A chemist's journey into Operational Technology Security" Mubina Mohammed and Graeme Clemens (AstraZeneca) "The use of automation for process development and understanding" Jordi Bures (University of Manchester) "Modern kinetic analyses for catalytic reactions" Jeremy Frey (University of Southampton) "GO FAIR, Go Far: Open data, Open science, Open innovation" Carole Mundell (University of Bath) "Spectral data management & analysis in astrophysics" Clemens Minnich (S-PACT) "Chemometrics - How to turn spectroscopic data into process knowledge" Many vendors also exhibited during the meeting, to enable users and vendors to share their thoughts on current issues and solutions alike. Vendors include ACD/Labs, Mestrelabs, Mettler-Toledo, TGK, Bruker, S-PACT and Advion. |
| Year(s) Of Engagement Activity | 2017 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/spectral_data_to_chemical_knowlegde/ |
| Description | Dial-a-Molecule. Synthetic Tools for the Exploration of 3D Pharmaceutical Space |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | The meeting was organised to learn about and discuss recent developments in the synthesis, analysis and biological evaluation of 3D molecules and their pharmaceutical importance, with the goal of inspiring new research in this area and fostering industry-academic collaborations. It was attended by over 90 delegates, including academics and students from all around the UK and Ireland, as well as several representatives from the pharmaceutical industry. Insightful talks were delivered by Professor Joe Sweeney (University of Huddersfield), Professor Rob Stockman (University of Nottingham), Dr Liam Cox (University of Birmingham), Professor Peter O'Brien (University of York), Professor Nigel Simpkins (University of Birmingham), Dr Susannah Coote (Lancaster University), Dr James Bull (Imperial College London), Dr Jeff St Denis (Astex), Dr Rob Young (GSK) and Dr Elaine O'Reilly (University of Nottingham). The meeting concluded with a fascinating keynote lecture given by Professor Erick Carreira (ETH Zurich) on his work on 'Recent Developments in Strategies and Tactics Towards Complex Secondary Metabolites as enabling tools for the Study of Natural Products Biology'. Other highlights included a demonstration of LLAMA (a free, open-access web interface able to predict the 3D shape and physical properties of molecules) by Dr James Firth, and a lively poster session, in which 23 PhD students, postdoctoral researchers and academics presented their work, with poster prizes awarded to Alexander Boddy (Imperial), Chris Jones (Queen Mary) Joan Mayol Llinas (Leeds) and Sam Griggs (York). |
| Year(s) Of Engagement Activity | 2017 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/3d_space/ |
| Description | Dial-a-Molecule/ROAR Reaction Kinetics Workshop |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Postgraduate students |
| Results and Impact | This 1 day workshop took place on Thursday, 31stJanuary, 2019 at the Molecular Science Research Hub (MSRH), Imperial College White City campus delivered by world leading expert, Prof. Donna Blackmond, Scripps Research Institute. The course offers an introduction to Reaction Progress Kinetic Analysis and provided a refresher on basic kinetic principles of organic reactions, especially catalytic reactions, focused on and applied to actual pharmaceutical examples, together with practical experience of extracting data from in-situIR spectroscopy and the use of IC kinetic software for processing. Topics covered included: • Introduction to Catalytic Reaction Rate Expressions • Application of Rate Expressions: Catalytic Hydrogenation as an Example Case • Reaction Progress Kinetic Analysis: Graphical Approach • Introduction to ReactIR • Use of IC kinetics software |
| Year(s) Of Engagement Activity | 2019 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/kinetics/ |
| Description | ECR Meeting 2019 - Supporting Synthesis and Self Assembly |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | Dial-a-Molecule and Directed Assembly Networks ECR Event Supporting Synthesis and Self Assembly 4-5 thJuly, 2019, University of York, Chemistry Department The event comprised stellar talks from Prof Marty Burke (Illinois) and Prof Adam Nelson (Leeds) as well as 2 minute flash talks presented by most of the ECRs by way of introduction - the topics varied widely reflecting the diversity of interests at the meeting and triggering the prospect of potential future collaborations. The formal sessions on day one were brought to a close by Dr Sarah Staniland (Sheffield) who gave a great talk on her work and included some personal reflections on career progression. The second day of the meeting focussed on academic-industry relationships, collaboration building and equality, diversity and inclusivity in chemistry. There were perspectives from Allan Watson (St. Andrews), Mei Lee (GSK) and Rob Davidson (Dr Reddy's) on different ways for academia and industry to work together. Anna Slater (Liverpool) gave an overview of a project she had been involved with through Vitae on the 'Experience of researchers taking maternity, paternity, adoption and/or parental leave', whilst Dave Smith talked through the key findings in the recently issued report 'Exploring the workplace for LGBT+ physical scientists' - both excellent pieces of evidence based research with recommendations on how the workplace needs to change. The sessions included lively discussions confirming the high level of interest of the topics to the audience. As well as scientific and career based discussions there was ample time for networking during the breaks and there was a wide ranging exhibition on show. |
| Year(s) Of Engagement Activity | 2019 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/ecr2019/ |
| Description | Early Career Researchers Meeting 2018 |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | The 3rdEarly Career Researchers' meeting for the Dial-a-Molecule and Directed Assembly EPSRC Grand Challenge networks took place on the 18th and 19th of June 2018, at the University of Strathclyde in Glasgow. The event was attended by approximately forty early career researchers from across the UK, including Belfast, Edinburgh, Glasgow, Leeds, London, Southampton, and York; early career attendees were typically in the early years of a first academic position, or working as postdoctoral researchers within a research group, and many elected to introduce themselves and their research through the medium of a two-minute flash presentation. There were three plenary speakers - Professor Ian Fairlamb (York), Dr Craig Jamieson (Strathclyde), and Professor Steve Goldup (Southampton) - who delivered honest, engaging, and informative presentations on their career paths and the science that they do aligned with the networks. As well as scientific discussions various organisations sent representatives to interact with early career researchers, explain what role they play in our communities, and how they can assist in delivering and disseminating high-quality research, and impact from that research. These included representatives from the Royal Society of Chemistry (RSC), the Engineering and Physical Sciences Research Council (EPSRC), ChemSinC, the Knowledge Transfer Network (KTN), and the new Rapid Online Analysis of Reactions (ROAR) facility based at Imperial College. |
| Year(s) Of Engagement Activity | 2018 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/ecr-meeting-2018/ |
| Description | Enabling Technologies for Synthesis Sandpit: Customisable Platforms for Thermo-, Electro-, Mechano- and Photo-chemistry |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | The aim of this event was to bring together cross-disciplinary cross-institutional teams to explore and develop new ideas and proposals in the highlighted areas. The workshop event consisted of speakers presenting the latest advanced techniques and technologies, and applications of these to synthesis. Speakers included: John Blacker (Reactor Design), Richard Brown (electrochemistry), Duncan Browne (mechanochemistry), Steve Hilton (3D printing), Daniele Leonori (photochemistry). The first day contained workshop/seminars on contemporary aspects of these techniques and the technologies. The evening consisted of flash poster presentations by the delegates on their own work in the area. The second morning was a facilitated workshop designed to coalesce ideas around future grant proposals. The aim of the event was to bring together teams and produce ideas suitable for EPSRC and equivalent proposals, whilst having access to experts to stress-test these ideas. |
| Year(s) Of Engagement Activity | 2017 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/enabling-technologies-synthesis/ |
| Description | Predicting Reaction Outcomes |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | This one day meeting, brought together chemists, engineers, computer scientists, mathematicians and data scientists from both academic and commercial backgrounds to provide a symposium and discussion forum looking at prediction of chemical reaction outcomes through data analysis and acquisition. In particular we aimed to identify the barriers currently preventing us from being able to predict reaction outcomes and to propose next steps. |
| Year(s) Of Engagement Activity | 2018 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/predictingreactionoutcomes/ |
| Description | Predictive scalability of processes in fine chemical and pharmaceutical manufacturing |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | The meeting took place on 29th November 2018 at the GSK site in Stevenage. Scale-up of complex organic and bio-catalytic processes remains a significant challenge due to low observability of state variables and, frequently, absence of mechanistic models of the reactions. A raft of new tools is emerging in this topic, relating to increased observability (new sensor methods), development of 'good-enough' black box models based on machine learning approaches, new kinetic methods, new methods in design of experiments and development of process models, as well as the increased understanding in physical organic chemistry and computer-aided molecular design. The meeting aimed at practitioners in the field, industrial and academic method developers as a networking and research agenda-setting event: what is the state-of-the-art, what research challenges are emerging as a result of developments in the neighbouring fields (robotics, AI, machine learning, systems biology, synthetic biology, high throughput analytics, novel business models for manufacturing, new feedstocks, etc), where is progress required? Talks included 'From Molecules to Reactors: Methods, Applications and Opportunities for Catalytic Process Design' (Dr Michail Stamatakis, UCL), 'Development and application of numerical simulation and optimization techniques for improved design and operation of complex reaction systems' (Dr Federico Galvanin, UCL), 'Rapid Development of Processes using Automated Flow Reactors and Machine Learning Algorithms' (Dr Richard Bourne, University of Leeds), 'A systematic function-based approach to optimal reactor and process design'(Prof. HannsjörgFreund, Friedrich-Alexander-Universität Erlangen-Nürnberg), 'Evaluating the Robustness of Industrial Scale Bioprocesses: Combining PAT and Scale-Down Bioreactors' (Prof. Peter Neubauer, Technische Universität Berlin), 'Application of process simulation for continuous processing' (Flavien Susanne, GSK), 'DynoChem prediction and simulation in process design and scale-up' (Drs Colm Cotter and Lucie Miller Potucka, AstraZeneca). The meeting concluded with a discussion session to identify future opportunities and challenges. |
| Year(s) Of Engagement Activity | 2018 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/process_understanding/ |
| Description | Promoting Accessibility of Enabling Technologies and Low-Cost Automation |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | The Meeting took place on Friday, 18th January, 2019 at the Materials Innovation Factory, University of Liverpool The aim was to provided researchers with information and ideas about how automation can be applied to different areas of research and elucidated options for seeking access to equipment in order to apply it to their own research ideas. There was a range of talks from both academia and industry presenting case-studies of the use of an enabling technology in their research, such as high-throughput synthesis platforms, flow reactors, or low-cost automation, to promote their use and demonstrate the benefits of embracing automation in research. The use of High-Throughput Automation Platforms was described by Andy Cooper (University of Liverpool), Varinder Aggarwal (University of Bristol), and Ben Deadman (Rapid on-line analysis of reactions facility, Imperial College London). Examples of Flow Chemistry and low-cost automation were described by Lee Cronin, University of Glasgow, Matthew O'Brien, Keele University, Nick Warren, University of Leeds, and Anna Slater, University of Liverpool. |
| Year(s) Of Engagement Activity | 2019 |
| URL | http://generic.wordpress.soton.ac.uk/dial-a-molecule/promaccess/ |
| Description | presentaion by coordinator (G Smith) at Leverhulme Research Centre Materials Discovery Symposium on 3-4 September 2018. |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | Highlighting Dial-a-Molecule, and particularly the need for automated experimentation principally to users and potential users of the Leverhulme Research Centre for Materials Discovery. |
| Year(s) Of Engagement Activity | 2018 |