LABORATORY INFORMATICS g


own compounds and assays, but we also believe that there is a treasure trove of hard to find, but valuable data in the public domain,’ Segall notes, and that is where the expertise of Medicines Discovery Catapult – a national, Innovate UK-funded scientific gateway to specialist expertise – comes into its own. ‘Medicines Discovery Catapult is using


natural language processing machine learning technology to find published papers with relevant data on compounds and structures, which can be used to build on what we already have and what our customers have, and provide a really rich source of data for these more hard-to-find targets,’ Segall added. ‘We have presented some early proof-of-concept work demonstrating that by combining all this expertise we can significantly outperform conventional QSAR approaches.’ The new machine learning technology


for ADMET is also versatile, and can be tuned to select results with a greater or lesser confidence. ‘We can ask the algorithm to present only those activities that can be predicted with the highest confidence. It’ll be a smaller number of results, but the data will be more reliable for decision making,’ stated Segall. Turn the predictive power dial the other


way and it is possible to take a bit more risk and include some more uncertain data, but you then get a bigger pot of compounds for selection. ‘Putting it all together, we believe this platform will be able to make more accurate predictions across a broader range of important endpoints for making decisions, and also learn in real time from new data as it is generated by our customers or in the literature,’ said Segall. While the pharmaceutical industry is


embracing deep learning technologies for discovery and development, a tendency towards ‘overhype’ of some research means that not all algorithms are created equal, Segall adds. ‘One needs to be very careful, because


“The companies who we are working with have big datasets of their own, based on their own compounds and assays, but we also believe that there is a treasure trove of hard to find, but valuable data in the public domain”


22 Scientific Computing World October/November 2018


in the drug optimisation space, for example, some of the new methods emerging are really more like conventional QSAR machine-learning based methods rebranded through good marketing. But the industry needs much more than just the same old stuff in a different box of tricks,’ said Segall. This new platform is designed to


represent the next step forward in being able to answer questions intuitively using the data already available, however fragmented. ‘Our goal is to give scientists the power to help make informed decisions on which compounds they can test, but also provide insight into the most appropriate assays for the next round of screening, to move promising compounds forwards towards optimised clinical candidates,’ Segall added.


Rapid detection Genedata is poised to launch an AI-based phenotypic image analysis software, Imagence, which has been developed to automate the workflows and analysis of otherwise massively time-consuming high-content screening data. Typically, this analysis is carried out during pharmaceutical R&D to evaluate the


effects of potential drug compounds on individual cells. While traditional high content screening


(HCS) methods rely on teams of scientists laboriously setting up the image analysis to identify and quantify compound-related phenotypic changes, Imagence takes this processing time down to literally seconds, explains Professor Stephan Steigele, head of science at Genedata. ‘The software uses machine learning


technology to understand, and then rapidly detect and extract cellular phenotypes that provide quantitative insight into the effects of an individual compound,’ he states. ‘Importantly, the software can be applied to a wide range of phenotypic assay formats and workflows. For the pharma industry, this means it is now possible to conduct massive screens involving millions of different chemical substances, in microtiter plate format, and generate far more consistent, reproducible and high-quality data for the discovery workflow than with classical computer vision,’ stated Steigele. So, traditional HCS approaches tie up


weeks of time as technicians, IT scientists and biologists must define and set up the screens and the image analysis


@scwmagazine | www.scientific-computing.com


Tex vector/Shutterstock.com


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36