SLAS Europe

June 27–29, 2018
Brussels, Belgium

Meet Genedata experts at booth #312 at SLAS Europe 2018 in Brussels, Belgium.

To get more information about Genedata Screener, please contact screener@genedata.com.

Recommended Oral Presentation

Scaling-Up Drug Discovery in the Fourth Dimension

Claire McWhirter, AstraZeneca, Cambridge, UK

Track: Late-Stage Discovery - Transforming the Lead to Candidate Process Through Innovation
Friday, June 29 | 10:00 - 10:30 | Copper Hall

Kinetic and mechanistic in-vitro characterisation of compounds yields critical predictive information on their potency in cellular assays, DMPK profile, and in-vivo performance. Thus, the ability to characterise more compounds this way in early SAR screening holds substantial benefit. Scalable instrumentation and lab automation to this end is available now. However, the complex data analysis easily takes a day per experiment, severely limiting throughput.
This talk outlines recent progress in developing plate based kinetic techniques along with advances in data analysis. A new software solution reduces analysis times ten-fold, while ensuring scalable, consistent, reliable processing of biophysical and mechanistic assays and direct publishing of results on a corporate level for effective ranking of lead series. With this new approach, we aim to dramatically scale up compound characterization embedded in a streamlined and modern pharma research workflow.

 

Compound Combination Special Interest Group (SIG)

Oliver Leven, chair

Wednesday, June 27 | 16:45 - 17:45 |Studio 312

The mission of the SLAS Compound Combination Screening SIG is to create a knowledge-sharing forum for screening practitioners active in the field of compound combinations. As such, the goal is to mature the field of compound combination screening, aimed at better science that accelerates the pace of drug discovery.

Development of a high-throughput cell line combination screening platform
Speaker: Lily-Xiaoyan Shi, Merck Biopharma

Combination therapies that improve treatment response, reduce adverse events and development of resistance have become the key for the treatment of cancer. Consequently, in-vitro drug Combination Screening has become important in oncology drug development. However, despite the recent advances in technology and methodology, combination screening remains a challenge due to its complexities in experimental designs, high data volume, data analysis and interpretation. Here we will share our experience in the development of a HTS combination screening platform. Over the last two years, we have set up an efficient and flexible high throughput combination process that allows us to run both large compound libraries screenings and small targeted screenings in multiple cancer cell lines. An essential step within this process is to efficiently analyze the data and feed the results into our corporate database.

 

HCS/HCA Data and Informatics Special Interest Group (SIG)

Daniel Siegismund, chair

Thursday, June 28 | 17:45 - 18:45 | Studio 312

Raising the bar in machine learning: Opportunities and Limitations of Deep Learning

Machine learning has been one of the big trends in HCS over the last few years, to the point where the application of such methods has become a commodity in the quantification of complex phenotypes. Deep learning has pushed this trend to a new level.  It has been shown to outperform existing machine learning methods in a range of applications and it eliminates the need to extract handcrafted features prior to the classification task.
This SIG will discuss the pros and cons of deep learning-based approaches, especially compared to classical machine learning approaches based on handcrafted features. The session will be split into three presentations, each given by experts in the fields, followed by a discussion with the audience.

Speakers: Peter Horvath, PhD and Emmanuel Gustin, PhD

Partnering to Close the Screening Loop: From Sample Logistics to Automation to Analysis and Back
Jack Dawson, Director of Applications, HighRes Biosolutions
Oliver Leven, Head of EU Business - Screener Business Unit, Genedata
Toby Winchester, Automation Product Manager, Titian Software

Workshop | Thursday, June 28 | 13:15 - 14:15 | Studio 311

The drug discovery research process consists of iterations between designing and planning, testing, and analyzing. HighRes Biosolutions, Titian, and Genedata together present an integrated set of products to support this iterative lifecycle. The approach offers a seamless workflow from sample logistics to automation to analysis and back, effortlessly moving information to where it needs to be and ultimately providing researchers with the time and freedom to focus on their science.
 
This tutorial demonstrates how a scientist designs and orders assay plates in Titian’s Mosaic, accepts and runs the fully automated screening process from within Cellario, and triggers the automated data analysis and approval in Genedata Screener. The overall workflow and its interfaces will be illustrated as well as the relevant interactions with the individual software components.

 

Recommended Poster Presentation

Illuminating the “Black Box” – Biological Discovery with Deep Learning   
Daniel Siegismund, Matthias Fassler, Marusa Kustec, Stephan Heyse and Stephan Steigele
Genedata AG, Basel, Switzerland
(Poster Number: P114)

In this poster we introduce a new approach – making the experimentalist “see” the findings of a deep learning network. We show methods to visualize and reason about the distinguishing aspects of cell images (morphologies, localized intensity differences …), e.g. separating control groups from each other. We illustrate exploration and validation of the classification process, shedding light into the “black box”. We also discuss how this approach may support new biological discoveries: Applied artificial intelligence can automatically identify new phenotypes in a truly unbiased way to catalyze R&D projects dealing with complex biology. 

[1] Kraus OZ, Grys BT, Ba J, Chong Y, Frey BJ, Boone C, & Andrews BJ (2017). Automated analysis of high‐content microscopy data with deep learning. Molecular systems biology, 13(4), 924.
[2] Godinez WJ, Hossain I, Lazic SE, Davies JW, & Zhang X (2017). A multi-scale convolutional neural network for phenotyping high-content cellular images. Bioinformatics, 33(13), 2010-2019.

 

Scaling-Up Drug Discovery in the Fourth Dimension
Claire McWhirter1, Artur Galimov3, Alexander Mehrle3, Geoff Holdgate2, Alexey Dubrovskiy3, Stephan Heyse3 and Stephan Steigele3
1Mechanistic Biology and Profiling, Discovery Sciences, IMED Biotech unit, AstraZeneca, Cambridge; 2Biophysics, Discovery Sciences, IMED Biotech unit, AstraZeneca, Cambridge; 3Genedata, Base
l
(Poster Number: P056)

Kinetic and mechanistic in-vitro characterisation of compounds yields critical predictive information on their potency in cellular assays, DMPK profile, and in-vivo performance. Thus, the ability to characterise more compounds this way in early SAR screening holds substantial benefit. Scalable instrumentation and lab automation to this end is available now. However, the complex data analysis easily takes a day per experiment, severely limiting throughput.

This poster outlines recent advances in data analysis to support the recent progress in developing plate based kinetic techniques along. A new software solution reduces analysis times ten-fold, while ensuring scalable, consistent, reliable processing of biophysical and mechanistic assays and direct publishing of results on a corporate level for effective ranking of lead series. With this new approach, we aim to dramatically scale up compound characterization embedded in a streamlined and modern pharma research workflow.

 

Kinetic profiling of protease inhibitors under different assay conditions
Cecilia Kankkonen
AstraZeneca, Moelndal
(Poster Number: P059)

Fibroblast activation protein alpha (FAPa) is an endopeptidase sharing 51% identity to DPPIV. FAPa is highly expressed in activated myofibroblasts in cancer and fibrotic diseases such as liver cirrhosis.  

Understanding the mechanism of action of the target enzyme can help to identify different types of inhibitors.

Here we show an example of mechanistic enzymology studies for FAPa where residence time and inactivation rate constant were determined and compared with data from Surface Plasmon Resonance experiments. Initial kinetic rates and jump dilution experiments provided insight in the kinetics of compound binding, but also allowed for discrimination between reversible and irreversible inhibitors. We also will show how detergent can have a major impact on the substrate Km as well as the kinetic binding profile for compounds. Examples of data analysis using the new Genedata mechanistic analysis module are shown.

Links:
slaseurope2018.org