18-19/11/2019 - AI3SD Network+ Conference - Holiday Inn Winchester & Winchester Science Centre

posted 26 Feb 2019, 08:31 by Samantha Kanza   [ updated 26 Feb 2019, 08:35 ]

More details coming soon. Save this date!

17/10/2019 - AI3SD & IoFT AI for Allergen Detection Workshop - SCI, London

posted 14 Feb 2019, 03:16 by Samantha Kanza   [ updated 26 Feb 2019, 08:43 ]

More details coming soon. Save this date!

12-13/09/2019 - AI3SD & Pistoia Hackathon - Wide Lane Southampton

posted 14 Feb 2019, 03:14 by Samantha Kanza   [ updated 26 Feb 2019, 08:42 ]

More details coming soon

11/09/2019 - AI3SD & ClearViewIP Workshop & Network+ TownMeeting - Wide Lane, Southampton

posted 14 Feb 2019, 03:13 by Samantha Kanza   [ updated 26 Feb 2019, 08:42 ]

More details coming soon

09-10/07/2019 - AI3SD & Dial-a-Molecule AI for Reaction Pathway Workshop - University of Cambridge

posted 14 Feb 2019, 03:12 by Samantha Kanza   [ updated 26 Feb 2019, 08:41 ]

More details coming soon

30/06/2019 - AI3SD AI 4 Good Workshop @ WebSci9 - Boston, US

posted 28 Jan 2019, 06:59 by Samantha Kanza   [ updated 26 Feb 2019, 08:48 ]


Description: 
This year the AI3SD Network+ (Artificial Intelligence and Augmented Intelligence for Automated Investigations for Scientific Discovery) will be running a workshop at the WebSci19 Conference in Boston. Artificial and Augmented Intelligence systems have the potential to make a real difference in the scientific discovery domain however this brings a new wealth of ethical and societal implications to consider with regards to this research (e.g. human enhancement, algorithmic biases, risk of detriment). This workshop looks to explore the ethical and societal issues centered around using intelligent technologies (Artificial Intelligence, Augmented Intelligence, Machine Learning, and in general Semantic Web Knowledge Technologies) to further scientific discovery, with a strong consideration of data ethics and algorithmic accountability. Advances in technology and software are rarely inherently bad in themselves, however that unfortunately does not preclude them from being subverted to ill intent by others; furthermore, as demonstrated by the examples above, even an unintentional lack of care towards ethical codes and algorithmic accountability can lead to societal and ethical implications of scientific discovery. It is our responsibility as researchers to consider these issues in our research; are we conducting studies ethically? What ethical codes can we put in place for scientific discovery research to mitigate against ethical and societal issues. These are really important issues, and they require an interdisciplinary focus between scientists, social scientists and technical experts in order to be comprehensively addressed.  

We are living through a data revolution, which will be as transformative of our society as the industrial revolution. Algorithms, and in particular, learning algorithms, are the engines of this revolution. ‘Intelligent’ algorithmic systems impact many areas of our personal and professional lives, making decisions based on prior ‘learned’ knowledge. The use of learning algorithms and has the potential to revolutionise scientific discoveries. However, these discoveries have the potential to be simultaneously beneficial and detrimental at the same time if they are not undertaken in a responsible and ethical manner. A few of these Major themes for this workshop are detailed below, although all other contributions surrounding the use of AI in Scientific Discovery are welcomed! 
  • AI in Drug Discovery & Heathcare – Intelligent technologies can be vastly useful in drug discovery and healthcare research as machine learning algorithms can be applied to vast linked datasets to make predictions that humans could not. However, this research may not apply to certain minority groups depending on the data used in the system. This may not be as a consequence of intentional algorithmic/data bias but equally is something that should have been addressed by ethical discussions at the start of a project, as “excluding minorities from healthcare research limits the ability to appropriately care for these population and skews the scientific understanding of disease” and indeed drugs to fight and cure these diseases.  
  • AI for Chemicals and Materials Discovery – Molecular compounds and materials underpin just about every aspect of our lives, from sustainable energy to healthcare. Society’s demands for enhanced performance is far outweighing our capability to discover materials that deliver it, so it is unsurprising that researchers are looking at using artificial intelligence and machine learning technologies to explore this space and speed up the discovery of new chemicals. However, do the algorithms to discover these chemicals and materials take into account whether they are environmentally friendly? Or whether certain chemicals could be psychoactive or extremely explosive? Whether materials could be used for ill intent? Does this mean chemicals/materials like this shouldn’t be investigated?
Key Dates:
  • Calls for Papers Opens: 22nd February 2019
  • Paper Submission: 1st April 2019
  • Notification of Acceptance: 16th April 2019 
  • Camera Ready Papers: 1st May 2019 
  • Workshop At WebSci19: 20th June 2019

01/05/2019 - AI3SD Semantics and Knowledge Learning for Chemical Design Workshop - Solent Conference Centre, Southampton

posted 3 Jan 2019, 08:42 by Samantha Kanza   [ updated 8 Feb 2019, 01:47 ]

Eventbrite Link: https://www.eventbrite.co.uk/e/ai3sd-semantics-and-knowledge-learning-for-chemical-design-workshop-tickets-55073920579 

Description: 
Designing chemicals, discovering new drugs, discovering materials and indeed all aspects of scientific discovery are all tasks that are highly data driven, and semantic web technologies are key to enabling researchers to deal with high levels of data in a useful and meaningful way. Semantic technologies facilitate representing data in a formal, structured, and interoperable way, and enable data to be reasoned over to infer potential relationships. In this workshop we seek to explore the ways in which semantic web technologies can be used to drive predictions in chemical design, including using Machine Learning and other AI techniques to exploit semantic links in knowledge graphs and linked datasets. There will be several keynote talks, an expert panel chaired by Professor Jeremy Frey, and a chance for breakout discussions. Lunch will be provided and the day will end with networking drinks.

Keynote Speakers
  • Dr Colin Bachelor - Colin Batchelor is a theoretical chemist by training, and now works at the Royal Society of Chemistry. He started as a technical editor before moving over to work on Project Prospect and ChemSpider before joining the Data Science team. He has published on natural language processing for chemistry, ontologies and cheminformatics.

  • Dr Age Chapman - Age is an Associate Professor of Computer Science at the University of Southampton, Co-Director of the ECS Centre for Health Technologies, and is part of the Web and Internet Science(WAIS) Research Group. Her research is in the area of database systems, focusing on using data appropriately and effectively. This involves solving problems that span the areas of databases, information discovery and retrieval, provenance and algorithmic accountability.

  • Dr Nicholas Gibbins - Nick is an Associate Professor in Computer Science at the University of Southampton and is part of the Web and Internet Science (WAIS) Research Group. His primary research interests are in the Semantic Web, Hypertext and Distributed Information Systems.

  • Professor Jonathan Goodman - Professor of Chemistry, Academic Dean; Director of Studies in Chemistry, University of Cambridge. Jonathan’s research focuses primarily on experimental and computational chemistry.

Keynote Abstracts
  • Challenging Chemistry: Solving Molecular Problems - Professor Jonathan Goodman: The power of AI is changing the way that chemistry is done, but chemistry is far from being solved. What is holding us back? Creating a new molecule or a necessary, but unknown, transformation is a demanding problem. Successes and surprises in predicting how molecules should behave can be used to try to discover the limits of our knowledge and the best ways to solve molecular problems.

  • Semantics vs. Statistics in Chemistry – Dr Colin Bachelor: Since the late 1990s, natural language processing (NLP) has seen a massive shift from high-precision, low-recall systems based on small sets of hand-written rules, to methods based on the statistical analysis of large corpora. The field of chemoinformatics, likewise, is dominated by statistical and machine-learning approaches. More recently deep learning methods have had surprising success in aspects of natural language processing, image processing and board games. Conversely, pharmaceutical companies have been engaging more and more with Semantic Web technologies, which are largely built around the sorts of hand-written systems that NLP has moved away from this century. In this talk I cover how we have applied both sorts of systems at the Royal Society of Chemistry and their strengths and weaknesses.

Programme
The programme for the day is as follows:
  • 10:00-10:30: Coffee & Registration
  • 10:30-11:00: Welcome Introduction – Professor Jeremy Frey
  • 11:00-11:30: Keynote – Dr Age Chapman
  • 11:30-12:00: Keynote - Dr Nick Gibbins
  • 12:00-12:30: Challenging Chemistry: Solving Molecular Problems – Professor Jonathan Goodman
  • 12:30-13:30: Lunch
  • 13:30-14:00: Semantics vs. Statistics in Chemistry - Dr Colin Bachelor
  • 14:00-15:00: Breakout Discussions
  • 15:00-15:30: Coffee & Report Back
  • 15:30-16:00: Keynote - TBC
  • 16:00-17:00: Panel Discussions chaired by Professor Jeremy Frey
  • 17:00-18:00: Drinks Reception
FAQs 

1. Who should attend? 
Anyone with an interest in Semantic Web Technologies (either in their own right or as a technology to be used in conjunction with Artificial Intelligence or Machine Learning), Scientific Discovery, Chemical Design, Knowledge learning for Scientific Discovery. We welcome members from academia, industry and government. We are always looking to grow our Network+ and bring in people with a wealth of experience in the many different subject areas that are needed so that we can form interdisciplinary partnerships and work together to further the field of Scientific Discovery.

2. What will I get out of it? 
You will be able to network with likeminded people who have research interests that complement yours. You will hear a range of thought-provoking talks about different aspects of semantics and knowledge learning for chemical design, and have the opportunity to both discuss this subject area with other members of the workshop, and address questions and raise areas of discussion to our expert panel. Members of the Network Executive Group will also be in attendance so you will be able to find out more about our Network and the opportunities we have available including funding opportunities and the types of events we will be running (e.g. workshops, conferences and hackathons). 

3. What are the aims of the workshop? 
This workshop is aiming to help the Network+ to drive progress in this area and facilitate collaboration by introducing people to make new interdisciplinary teams, and to produce new grant applications. To achieve this we may commission literature reviews, papers, or small scale investigations to test out new ideas. We welcome ideas and suggestions about how to go forward in this area and how best to achieve our aims.

19/03/2019 - AI3SD AI for Materials Discovery Workshop - University of Southampton

posted 3 Jan 2019, 08:27 by Samantha Kanza   [ updated 18 Feb 2019, 02:35 ]


Description 
Materials play a key role in modern society and the growing demands for functionality, selectivity, re-usability, efficiency and environmentally sustainable which all place huge demands on the chemistry.  The use of Computational Chemistry and more generally Artificial and Augmented Intelligence (AI) to predict structure and function provides new possibilities in using theory to drive innovation and discovery. 

Agenda 
  • 13:00-13:30 - Registration & Coffee 
  • 13:30-13:45 - Welcome from Professor Jeremy Frey 
  • 13:45-14:30 – Theoretical Studies of CO and CO2 Hydrogenation to Methanol and Conversion of Methanol to Olefins – Professor Felix Studt  
  • 14:30-15:15 - LAISER: Putting the AI in Laser - Dr Ben Mills 
  • 15:15-15:30 - Coffee Break 
  • 15:30-16:15 - Machine learning opportunities in prediction-led discovery of molecular materials – Professor Grame Day 
  • 16:15-17:00 - Potential Solutions to Mathematical Challenges for Solid Crystalline Materials – Dr Vitaliy Kurlin 
  • 17:00-17:45 – One million crystal structures: what can we learn? – Dr Angeles Pulido 
  • 17:45-18:00 - Wrap up & Conclusions
Keynote Speakers 
  • Professor Felix Studt - Felix is a Professor at the Institute of Chemical Technology and Polymer Chemistry (ITCP) and the Institute of Catalysis Research and Technology (IKFT) at the Karlsruhe Institute of Technology (KIT). His research interests are in theory guided materials discovery, electrochemical processes, synthesis gas conversion, CO2 reduction, and routes from biomass to chemicals. He has published over 70 papers which have had over 2900 citations, and has been invited to speak at many international conferences. He has also published a textbook on the  Fundamental Concepts in Heterogeneous Catalysis.  
  • Dr Ben Mills - Ben is a Senior Research Fellow and EPSRC Early Career Fellow at the Optoelectronics Research Centre at the University of Southampton, where he leads a research group focussed at the interface of laser machining and machine learning. Ben completed his PhD in high harmonic generation via ultrafast lasers at the University of Southampton in 2009, and then became the manager of the “FAST lab”, a femtosecond laser facility at the University. His background covers 10 years in ultrafast lasers and their application for high-precision materials processing. Current research interests also include laboratory automation and machine learning, in particular convolutional neural networks. 
  • Professor Graeme Day - Graeme is a Professor of Chemical Modelling at the University of Southampton. His research interests centre on the development and application of computational methods for understanding and predicting the structures and properties of molecular materials. An area of particular interest is crystal structure prediction, and its applications in structure determination, polymorph discovery and the design of materials with targeted properties. These research areas all stem from a fundamental interest in understanding and modelling intermolecular interactions. Graeme is the author or co-author of over 115 publications, including 5 book chapters. He serves on the advisory board for the Royal Society of Chemistry’s journal Molecular Systems Design and Engineering, is on the steering committee of the UK Materials Chemistry High End Computing Consortium and is a member of the EPSRC peer review college.
  • Dr Vitaliy Kurlin - Vitaliy is a Computer Scientist at the Materials Innovation Factory in Liverpool, where he facilitates the collaboration between Chemists and Computer Scientists. He was awarded the Marie Curie International Incoming Fellowship (2005-2007) and the EPSRC grant “Persistent Topological Structures in Noisy Images" (2011-2013). In 2014-2016 he has gained industrial experience through Knowledge Transfer Secondments in the Computer Vision group at Microsoft Research, Cambridge, UK. From 2018 he leads the Liverpool team on a £2.8M EPSRC 5-year grant “Application-Driven Topological Data Analysis” (with Oxford and Swansea). His research group includes one postdoc and five PhD students working on applications of topology and geometry to Materials Science, Computer Vision and Climate. 
  • Dr Angeles Pulido - Angeles is a Research and Application Scientist at the  Crystallographic Data Centre (CCDC) the Cambridge CDC, she is  part of the Pfizer Design Centre within the Materials Science team who apply computational techniques to study organic molecular crystals relevant to pharmaceutical industry, with especial interest in crystal structure prediction, materials stability and polymorphism. Angeles’ main research interest is in silico modelling of solids and the use of computational techniques to provide an atomistic view and a better understanding of thermodynamic, kinetic and spectroscopic features of crystalline organic and inorganic materials.
Keynote Abstracts 
  • Theoretical Studies of CO and CO2 Hydrogenation to Methanol and Conversion of Methanol to Olefins – Professor Felix Studt: The catalytic conversion of CO2 to fuels and chemicals is experiencing renewed interest and growth as it is seen as one of the cornerstone reactions in a future sustainable energy scenario. Methanol, which can also be produced from CO2, is also an important chemical building block as it can be converted to olefins, hydrocarbon and gasoline. Theoretical studies of the processes at the catalytic surfaces help to understand how these catalyst function on the atomic-scale. Here insight gained on the active site of methanol synthesis[1] as well as the selectivity for CO and CO2 hydrogenation[2] is used for the computational screening of new CO2 hydrogenation catalysts.[3] We also investigated the conversion of methanol to olefins in zeolite catalysts using a combination of ab initio/density functional theory and microkinetic/reactor modeling.[4,5] In addition we will show how theory can help establishing trends across different acid sites and various frameworks,[6-8] a finding that might serve as a guidance for the discovery of improved catalysts for the production of fuels and chemicals from methanol. 
  • LAISER: Putting the AI in Laser – Dr Ben Mills: Advances in lasers now allow the laser-based processing of almost any material. Innovation in this field is now becoming heavily focussed on making existing processing techniques more precise and efficient. A research area of particular current importance is therefore the development of real-time monitoring and feedback systems for laser machining, via visual inspection of the sample during machining. Convolutional neural networks (CNNs) offer the capability for image processing without the need for understanding the underlying physical processes, and hence offer an ideal solution for the monitoring of laser machining, which itself is not fully understood. In this talk, the application of CNNs for real-time monitoring and process control for laser machining will be discussed, along with the capability of CNNs for predicting the outcome of laser machining before the experiment occurs. In addition, an application of combining laser light with CNNs for real-time sensing of pollution particulates will be demonstrated. 
  • Machine learning opportunities in prediction-led discovery of molecular materials - Professor Graeme Day: Predictive computational approaches have developed rapidly as tools to accelerate the discovery of molecular materials with targeted properties. A challenge in developing the use of these approaches is the expense of both crystal structure prediction and property prediction, and the difficulty of interpreting the resulting energy-structure-function landscapes, which normally contain huge numbers of possible structures. The talk will discuss opportunities for developing machine learning approaches to improve the speed and reliability of computational predictions. 
  • Potential Solutions to Mathematical Challenges for Solid Crystalline Materials – Dr Vitaliy Kurlin: Abstract. Solid crystalline materials (briefly, crystals) can be modelled as periodic structures based on a geometric pattern that represents any chemical composition. One of the challenges in crystal structure prediction is to encode any crystal in a unique numerical form that is convenient to compare crystals and to search for new crystals with better properties. The talk will discuss continuous geometric invariants that will enable a more efficient search in the huge configuration space of all possible crystals. 
FAQ 

1. Who should attend? 
Anyone with an interest in Materials discovery, Artificial Intelligence, Machine Learning, Deep Learning, and particularly those looking to apply AI technologies to materials discovery. We welcome members from academia, industry and government. We are always looking to grow our Network+ and bring in people with a wealth of experience in the many different subject areas that are needed so that we can form interdisciplinary partnerships and work together to further the field of Scientific Discovery. 

2. What will I get out of it? 
You will be able to network with likeminded people who have research interests that complement yours. There will be several keynotes around the topics of AI for Materials Discovery to spark discussion and ideas. You will hear a range of thought-provoking talks about different aspects of using AI technologies in the area of materials discovery, and have the opportunity to both discuss this subject area with other members of the workshop and address questions to the speakers. Members of the Network Executive Group will also be in attendance so you will be able to find out more about our Network and the opportunities we have available including funding opportunities and the types of events we will be running (e.g. workshops, conferences and hackathons). 

3. What are the aims of the workshop? 
This workshop is aiming to help the Network+ to drive progress in this area and facilitate collaboration by introducing people to make new interdisciplinary teams, and to produce new grant applications. To achieve this we may commission literature reviews, papers, or small scale investigations to test out new ideas. We welcome ideas and suggestions about how to go forward in this area and how best to achieve our aims. 

4. What are the main themes of the workshop? 
Materials, Computational chemistry, Novel Mathematics, AI.

06/03/2019 - AI3SD & MDC AI in Drug Discovery and Drug Safety Workshop - Medicines Discovery Catapult, Cheshire

posted 3 Jan 2019, 07:58 by Samantha Kanza   [ updated 21 Mar 2019, 12:03 ]

Eventbrite Link: https://www.eventbrite.co.uk/e/ai-in-drug-discovery-and-drug-safety-workshop-tickets-51705581787 

Description: 
This is event is brought to you by AI3SD and Medicines Discovery CatapultDrug discovery is a complex and long-term scientific investigation involving interdisciplinary research methods coupled with large heterogeneous datasets. The research and data space in this area is vast, and we believe that the use of AI and machine learning technologies can help spur on advances in this domain. This workshop has been designed to draw together those with a keen interest in using AI and machine learning technologies in the domain of drug discovery, both to aid future drug discovery, and to help improve drug safety. At AI3SD we firmly believe that interdisciplinary collaboration is the key to many of these advances, and so welcome anyone working in the technical or scientific ends of this domain, as well as those already working in an interdisciplinary fashion. There will be keynote talks interspersed with general group discussions, and working groups around the key topics that arise. There will also be an opportunity to tour the labs at Medicines Discovery Catapult. Lunch will be provided and there will be plenty of time for networking, and the day will conclude with a prosecco reception.

Keynote Speakers:
Alongside the discussions and lab tours there will some keynote talks about different aspects of drug discovery: 
  • Professor John Overington, Chief Informatics Officer, Medicines Discovery Catapult. John leads the development and application of informatics approaches to promote and support innovative, fast-to-patient drug discovery in the UK through collaborative projects across the applied R&D community.
  • Professor Val Gillet - Val is a Professor of Chemoinformatics at the University of Sheffield where she heads the Chemoinformatics Research Group. Her research focuses on the development and validation of chemoinformatics methods, especially for drug discovery. She has expertise in machine learning, evolutionary algorithms and the development of novel methods for molecular representation and applying these to applications such as de novo design and virtual screening. She has collaborated with many of the major pharmaceutical companies and specialist chemcoinformatics software companies.
  • Dr Willem van Hoorn, Chief Decision Scientist, ExScientia. Willem gained a PhD in computational chemistry in the group of David Reinhoudt at the University of Twente followed by a postdoc at Yale. He subsequently spent a decade at Pfizer focusing on computational techniques for HTS triage and combinatorial library design. This was followed by a position as senior solutions consultant at Accelrys assisting a range of clients from small biotech to big pharma.
  • Dr Nicola Richmond – Nicola is the Director of Artificial Intelligence and Machine Learning at GlaxoSmithKline. Her research focuses on discovering innovative ways of deriving insights from data to advance drug discovery and development.
Keynote Abstracts:
  • Using Machine Learning to Drive Reaction Based De Novo Design - Professor Val Gillet: The de novo design of novel drug candidate has been a topic of considerable interest since the 1990s. The main challenges in de novo design arise from the astronomical number of drug-like molecules that could exist and the difficulties associated with designing scoring functions to navigate this space. The recent resurgence of interest in de novo design can be attributed to the application of deep learning methods that typically operate on SMILES strings. While these approaches have been shown to be effective in generating valid SMILES, they are limited in the extent to which they can account for synthetically accessibility. We have been working on reaction-based de novo design for a number of years. Our approach takes explicit account of synthetic accessibility since the transformations that are applied to molecules are based on rules derived from real reactions. The rules are encoded as reaction vectors and are derived automatically from reaction databases. The availability of large public datasets of reactions provides a rich source of reactions for synthetically accessible de novo design. Here we will describe how we are using machine learning to select the most promising reactions for reaction-based de novo design. 
  • Re-energising Small Molecule Drug Discovery – Dr Willem van HoornThe optimisation trajectory of hit to lead to candidate is the most expensive part of drug discovery. Exscientia’s drug discovery platform brings that cost down significantly by combining the strengths of AI compound design and human strategic thinking into the Centaur ChemistTM. A high level overview of the technology is presented and results are shown from successful collaborations that resulted in the delivery of clinical candidates in less than a year.
  • Understanding the holes in the metabolome - Dr Nicola Richmond: The metabolome refers to the complete set of both endogenous and exogenous small molecule metabolites that are either produced naturally as a bi-product of a biological process or as a result of the external environment. Quantifying changes in the metabolome can help diagnose disease, understand disease mechanisms, identify novel drug targets and understand drug safety and efficacy. As such, metabolomics is now widely used in the pharmaceutical industry throughout the drug discovery and development process. The annotated human metabolome now stands at over 350K metabolites and 25K pathways. It is therefore unsurprising that analysing metabolomics data presents a major challenge. The current gold standard approach is highly subjective and does not account for pathway-level, structural information. Hypotheses tend to be established a priori and validated through manual navigation of data rather than letting the data speak. At GSK, we have established a fully automated, data-driven approach to analysing metabolomics data using concepts from topological data analyses. Our analysis pipeline provides bench scientists with an automated approach for validating their hypotheses, allows data scientist, with no understanding of biology, to generate meaningful hypotheses and potentially fills gaps in our understanding of the metabolite.
Programme: 
  • 10:00 - 10:30: Coffee & Registration
  • 10:30 - 10:45: Introduction with Professor John Overington
  • 10:45 - 11:15: Using Machine Learning to Drive Reaction Based De Novo Design - Professor Val Gillet
  • 11:15 - 12:30: Initial Discussions
  • 12:30 - 13:00: Lunch
  • 13:00 - 13:30: Re-energising Small Molecule Drug Discovery – Dr Willem van Hoorn
  • 13:30 - 15:00: Working Group Discussions and Lab Tours
  • 15:00 - 15:15: Coffee
  • 15:15 - 15:45: Report and Action Plan
  • 15:45 - 16:15: Understanding the holes in the metabolome - Dr Nicola Richmond
  • 16:15 - 17:30: Prosecco reception
FAQS:

1. Who should attend?
Anyone with an interest in Drug Discovery and Drug Safety, and Artificial Intelligence, Machine Learning or Data Science and how these methods can be applied to Drug Discovery/Safety. We welcome members from academia, industry and government. We are always looking to grow our Network+ and bring in people with a wealth of experience in the many different subject areas that are needed so that we can form interdisciplinary partnerships and work together to further the field of Scientific Discovery.

2. What will I get out of it?
You will be able to network with likeminded people who have research interests that complement yours. You will hear a range of thought-provoking talks about different aspects of using AI and Machine Learning in Drug Discovery and Drug Safety. There will also be plenty of time to network and discuss this subject area with other members of the workshop, in addition to being able to take a tour of the Medicines Discovery Catapult labs. Members of the Network Executive Group and Advisory Board will also be in attendance so you will be able to find out more about our Network and the opportunities we have available including funding opportunities and the types of events we will be running (e.g. workshops, conferences and hackathons).

3. What are the aims of the workshop?
This workshop is aiming to help AI3SD and MDC drive progress in this area and facilitate collaboration by introducing people to make new interdisciplinary teams, and to produce new grant applications. To achieve this we may commission literature reviews, papers, or small scale investigations to test out new ideas. We welcome ideas and suggestions about how to go forward in this area and how best to achieve our aims.

06/02/2019 - AI3SD Molecules Graphs & AI Workshop - Ageas Bowl, Southampton

posted 3 Jan 2019, 07:56 by Samantha Kanza   [ updated 28 Jan 2019, 07:22 ]

Eventbrite Link: https://www.eventbrite.co.uk/e/molecules-graphs-ai-workshop-tickets-55016184890 

Description: 
The representation of molecules as connected graphs and the application of graph theory has been very useful in defining aspects of molecular structure, molecular energy levels and identifying unique topology features. In the workshop we seek to explore the ways in which molecular graphs can be used to drive property and other predictions using Machine Learning and other AI techniques. All ideas welcome – come and discuss and debate and come up with new plans! There will be two keynote talks, and a chance for participants to present their areas of research and ideas. Lunch will be provided, and the day will end with networking drinks.

Keynote Speakers
There will be three keynote talks from experts in this field.
  • Professor Mahesan Niranjan - Professor of Electronics and Computer Science at the University of Southampton & AI3SD Co-Investigator. Nirajnan works in the area of machine learning, and his research interests are in the algorithmic and applied aspects of the subject. He has worked on a range of applications of machine learning and neural networks including speech and language processing, computer vision and computational finance. Currently, the major focus of his research is in computational biology. Some of his work (e.g. the SARSA algorithm in Reinforcement Learning) have been fairly influential in the field. He has held several research grants from the Research Councils in the UK, and the European Union. Currently, his main focus is on architectures and algorithms for Deep Learning and inference problems that arise in computational biology.
  • Professor Sophia Yaliraki - Sophia is a Professor of Theorerical Chemistry at Imperial College London. Sophia is a Professor of Theorerical Chemistry at Imperial College London. Her research interests are in the theory of self-assembly in biology and molecular electronic devices and coarse-graining and model reduction techniques in dynamics.
  • Professor Patrick Fowler - Patrick has been a Professor of Theoretical Chemistry at the University of Sheffield since 2005. Prior to that he had worked at both the University of Durham as a Senior Demonstrator, the University of Cambridge as a Postdoctoral Research Fellow and the University of Exeter where he became a professor. His research focuses on molecular properties, ring currents, aromaticity, fullerenes, molecular electronic devices, symmetry and discrete mathematics in chemistry, and he has published extensively in these areas. He was also elected a Fellow of the Royal Society in 2012.
Keynote Abstracts
  • Inference from Outliers – Professor Mahesan Niranjan: Classic machine learning is largely about classification and regression problems. However, many practical problems of interest in genomics, condition monitoring, medical diagnostics and security are better posed as problems of detecting novelty. In this talk, I will describe two applications of extracting useful information from novel data, in problems relating to modelling cellular protein concentrations and the solubility of synthetic chemical molecules. The algorithmic framework poses a robust support vector regression problem and the resulting non-convex optimisation problem is solved using a difference-of-convex formalism. (Part of this work is supported by grant EP/N014189/1, "Joining the Dots: From Data to Insight" from the EPSRC).
  • Unsupervised, multiscale learning through atomistic graphs: From molecules to systems - Professor Sophia Yaliraki: We have derived an all-scale graph partitioning approach that preserves atomistic physico-chemical detail and by using diffusive processes on the graph (both on the node and the edge space), we have shown that we can obtain the behaviour of biomolecules and biomolecular assemblies at different timescales without the need of any reparametrisation or a priori selection of relevant timescales. The approach is computationally efficient and general and can be applied to molecules, molecular assemblies as well as data. We will showcase the theory with examples from predictions and experimental verification of mutations that control protein dynamics at different scales (AdK), prediction of allosteric sites for drug design and communication and signalling in multimers and assemblies (ATCase, Rubisco). Finally, the application of this unsupervised learning approach to trajectories and free text will be briefly discussed.
  • Source-and-sink models for molecular conduction - Professor Patrick Fowler: This talk describes recent progress in Sheffield in describing ballistic molecular conduction with the Ernzerhof source-and-sink-potential (SSP) model. SSP gives a broad classification of conduction behaviour at the graph theoretical level. We have been able to derive selection rules, classifications and intuitive descriptions that remain useful at higher levels of theory. This talk is based on joint work with Barry Pickup (Sheffield), Irene Sciriha (Malta) and Martha Borg (Sheffield).
Programme: 
  • 10:00-10:30: Coffee & Registration 
  • 10:30-10:45: Welcome Introduction – Professor Jeremy Frey 
  • 10:45-11:15: Unsupervised, multiscale learning through atomistic graphs: From molecules to systems - Professor Sophia Yaliraki
  • 11:15-12:00: Inference from Outliers - Professor Mahesan Niranjan  Presentations from Participants to initiate discussions
  • 12:00-12:45: Presentations from Participants to initiate discussions
  • 12:45-13:45: Lunch 
  • 13:45-14:15: Source-and-sink models for molecular conduction - Professor Patrick Fowler
  • 14:15-14:45: Initial Discussions to form Working Group Topics 
  • 14:45-15:15: Coffee 
  • 15:15-16:30: Working Group Discussions 
  • 16:30-17:00: Report back and form Action Plan 
  • 17:00-18:00: Drinks reception
FAQs:

1. Who should attend? 
Anyone with an interest in Molecules, Graphs, Artificial Intelligence, Machine Learning, Deep Learning. We welcome members from academia, industry and government. We are always looking to grow our Network+ and bring in people with a wealth of experience in the many different subject areas that are needed so that we can form interdisciplinary partnerships and work together to further the field of Scientific Discovery.  

2. What will I get out of it? 
You will be able to network with likeminded people who have research interests that complement yours. There will be two keynotes around the topics of Molecules, Graphs and AI to spark discussion and ideas. There will be an opportunity to present your own research interests/areas of expertise briefly, with plenty of opportunity to have general discussions and some specific topic based discussions in smaller groups. Members of the Network Executive Group will also be in attendance so you will be able to find out more about our Network and the opportunities we have available including funding opportunities and the types of events we will be running (e.g. workshops, conferences and hackathons).  

3. What are the aims of the workshop? 
This workshop is aiming to help the Network+ to drive progress in this area and facilitate collaboration by introducing people to make new interdisciplinary teams, and to produce new grant applications. To achieve this we may commission literature reviews, papers, or small scale investigations to test out new ideas. We welcome ideas and suggestions about how to go forward in this area and how best to achieve our aims.  

4. What are the main themes of the workshop? 
Molecules are often represented by connected graphs showing in some sense the bonding between the atoms.  These graphs can serve as the input in the quantum chemistry packages to determine the 3D molecular geometry and with high level and time-consuming calculations obtain the electron density and electric fields surrounding the molecules and with even greater difficulty the interactions between molecules obtained and used as the basis for simulation for collections of molecules.  In this workshop we wish to explore how the graphs themselves can be used to generate molecular properties, predict drug activity, suggest crystal structures, without going through the QM route.  Can the graphs be the input to Machine Learning models?  What fundamental properties of the graphs directly relate to molecular behaviour? How do these relate to the topology of the systems?  Can we predict drug solubility, drug binding or the nature of molecular assemblies and crystals structures?  For example graph theory predictions of Huckel Molecular orbital energies give very fundamental predictions about the nature of some molecules. The graph approach is very useful when considering similarity between molecules and the transformations between molecules  and we seek to explore how AAI can play a role in these areas too.

1-10 of 12