This seminar forms part of the AI3SD Online Seminar Series that will run across the winter (from November 2020 to April 2021). This seminar will be run via zoom, when you register on Eventbrite you will receive a zoom registration email alongside your standard Eventbrite registration email. Where speakers have given permission to be recorded, their talks will be made available on our AI3SD YouTube Channel. The theme for this seminar is Enhancing Experiments through Machine Learning.
This seminar was the second of ten in our AI3SD Winter Seminar Series. This seminar was hosted online via a zoom webinar. The event was themed around Enhancing Experiments through…
This seminar forms part of the AI3SD Online Seminar Series that will run across the winter (from November 2020 to April 2021). This seminar will be run via zoom, when you register on Eventbrite you will receive a zoom registration email alongside your standard Eventbrite registration email. Where speakers have given permission to be recorded, their talks will be made available on our AI3SD YouTube Channel. The theme for this seminar is Robots, AI and NLP in Drug Discovery.
There is now broad recognition within the scientific community that the ongoing deluge of scientific data is fundamentally transforming academic research. Turing Award winner Jim Gray referred to this revolution as “The Fourth Paradigm: Data Intensive Scientific Discovery’. Researchers now need tools and technologies to manipulate, analyze, visualize, and manage vast amounts of research data. This talk will begin by reviewing the challenges posed by the explosive growth of experimental and observational data generated by large-scale facilities such as the Diamond Synchrotron and the CryoEM Facilities at the Rutherford Appleton Laboratory. Increasingly, scientists are beginning to use sophisticated machine learning and other AI technologies both to automate parts of the data pipeline and also to find new scientific discoveries in the deluge of experimental data. In particular, ‘Deep Learning’ neural networks have already transformed several areas of computer science and research scientists are now exploring their use in analyzing their ‘Big Scientific Data’. The talk concludes with a vision of how this ‘AI for Science’ agenda can be truly transformative for experimental scientific discovery.
Since the 1980’s the development of novel antibiotics has dramatically reduced. This, combined with the ever-increasing prevalence of antibiotic resistance in bacteria, means that some bacterial strains have now been identified that are resistant to treatment with all known classes of antibiotic currently available. Supramolecular Self-associating Amphiphiles (SSAs) are a novel class of amphiphilic salts that contain an uneven number of covalently linked hydrogen bond donating and accepting groups, meaning that they are ‘frustrated’ in nature. The hydrogen-bonded, self-associative properties for members of this class of over 70 compounds synthesised to date have been extensively studied in the gas phase, solution state, solid state and in silico. Through these studies we have shown correlations between certain physicochemical properties that maybe predicted by simple, low-level, high-throughput, easily accessible computational modelling. In addition, members from this class of compound have been shown to kill a variety of different bacteria, including those with known antibiotic resistance (e.g. Methicillin Resistant Staphylococcus aureus (MRSA)). These initial studies have highlighted within the supramolecular chemistry community a vast amount of experimental data, not yet accessed by AI/machine learning. Could data sets such as these be the next targets of interest for this community? Is there room for a consortium or community led approach to solving predictive modelling within this branch of chemistry.
Presenting the paper: On the Basis of Brain: Neural–Network–Inspired Changes in General Purpose Chips. In this paper, we disentangle the changes that the rise of Artificial Intelligence Technologies (AITs) is inducing in the semiconductor industry. The prevailing von Neumann architecture at the core of the established “intensive” technological trajectory of chip production is currently challenged by the rising difficulty to improve product performance over a growing set of computation tasks. In particular, the challenge is exacerbated by the increasing success of Artificial Neural Networks (ANNs) in application to a set of tasks barely tractable for classical programs. The inefficiency of the von Neumann architecture in the execution of ANN-based solutions opens room for competition and pushes for an adequate response from hardware producers in the form of exploration of new chip architectures and designs. Based on an historical overview of the industry and on collected data, we identify three characteristics of a chip — (i) computing power, (ii) heterogeneity of computation, and (iii) energy efficiency — as focal points of demand interest and simultaneously as directions of product improvement for the semiconductor industry players and consolidate them into a techno– economic trilemma. Pooling together the trilemma and an analysis of the economic forces at work, we construct a simple model formalising the mechanism of demand distribution in the semiconductor industry, stressing in particular the role of its supporting services, the software domain. We conclude deriving two possible scenarios for chip evolution: (i) the emergence of a new dominant design in the form of a “platform chip” comprising heterogeneous cores; (ii) the fragmentation of the semiconductor industry into submarkets with dedicated chips. The convergence toward one of the proposed scenarios is conditional on (i) technological progress along the trilemma’s edges, (ii) advances in the software domain and its compatibility with hardware, (iii) the amount of tasks successfully addressed by this software, (iv) market structure and dynamics.
he concept of in silico molecular design goes back decades and has a long history of published approaches using many different algorithms and models. Major challenges involved in de novo molecular design are manifold, including identifying appropriate molecular representations for optimisation, scoring designed molecules against multiple modelled endpoints, and objectively quantifying synthetic feasibility of the designed structures. Recently, multiobjective de novo design, more recently referred to as generative chemistry, has had a resurgence of interest. This renaissance has highlighted a step-change in successful applications of such methods. This presentation will review the development of de novo design methods over the years including the author’s original work in this area from the early 2000s, to recent approaches that show great promise. Through this review, improvements in important components of de novo design, including machine learning model predictions and automated synthesis planning, will also be presented.
Professor Charlotte Deane from the University of Oxford speaks about some of the work her research group have done on Machine Learning for Early Stage Drug Discovery to give a flavour of the different kinds of approaches they have been looking at. These run from predicting whether molecules will bind or not bind to a given protein target, to trying to remove biases from that kind of work, to finally how do we generate novel molecules in the protein binding sites.
https://www.youtube.com/watch?v=GY0myVuhrCo&t=14s&ab_channel=AI4ScientificDiscovery Abstract: Professor Charlotte Deane from the University of Oxford speaks about some of the work her research group have done on Machine Learning for Early Stage Drug Discovery to…
https://www.youtube.com/watch?v=Gc7MGnQ4mEk&ab_channel=AI4ScientificDiscovery Abstract: Symbolic AI, sometimes referred to as Good Old-fashioned AI, has its roots in the earliest days of the AI project. It seeks to represent reasoning using explicit data…