Efficient SAR analysis and reporting for rapid compound optimization: A Discngine-Bayer success story
“The solution developed with Discngine allows us to analyze and discuss multi-parameter SAR results in a data-driven, visual way via an easy-to-use tool”
Is efficient, multi-parameter Structure-Activity Relationship (SAR) analysis still a challenge for discovery chemists? It might very well be.
Compound optimization projects must deal with large volumes of bioactivity data. For thousands of compounds, performance in multiple biochemical and biological assays needs to be analyzed in parallel to inform the next compound designs. The use of outdated, siloed technology often makes this multi-objective SAR analysis slow and inefficient.
For the chemists at Bayer Crop Science, these challenges were all too familiar. Struggling with complex spreadsheets and time-consuming data management to analyze their compounds' performances, they were looking for a modern, digital solution. The goal was to allow chemists to streamline SAR analysis, reduce manual efforts, and enhance visualization to focus on what truly matters – compound optimization and decision–making – rather than struggling with data and technology.
That’s when they turned to Discngine. With our custom research services and a team of dual-skilled scientific and IT consultants, we helped Bayer Crop Science experts overcome inefficiencies in SAR analysis and SAR report generation.
In this article, you will discover the collaborative Discngine and Bayer success story, showcasing two parallel projects complementing one another: The initial Matched Molecular Pairs (MMPs) project, followed by the SAR Slides project. The article is based on insights from interviews with experts leading these projects, including:
Britta Nisius, Data Scientist Cheminformatics, and Janosch Achenbach, Data Scientist & Digital Transformation Lead, Bayer Crop Science, Germany - leaders of “MMPs project” which focused on developing tools for multi-parameter SAR analysis based on the MMPs methodology
David Bernier and Aurélie Mallinger, Lab Leaders, Bayer Crop Science France, who built upon the MMPs project to develop the SAR Slides application for generating high-quality SAR reports
Article highlights:
Introduction: Initial challenges
Towards a comprehensive tailored solution
Discngine impact
Real-world benefits on compound discovery projects
Introduction
Limited technology for multi-parameter SAR analysis and report generation
“MMPs” project needs – Systematic SAR analysis
Britta: Five years ago, we faced a significant challenge: there was no straightforward and visual way for chemists and data scientists to objectively perform multi-parameter SAR analysis. Most analyses relied on spreadsheets, which became increasingly cumbersome – dozens of compound columns and an ever-expanding list of parameters to consider, including critical factors like safety and sustainability. Even for experienced data scientists, this setup became overwhelming.
Janosch: On top of multi-dimensional analysis challenges, exchanging data and sharing findings with colleagues was very difficult. Everyone conducted their analysis differently, mostly starting from scratch each time. Even after successfully navigating the time-consuming analysis, another challenge remained – presenting the results in a proper context for colleagues to understand, which took additional time. Overall, the entire analysis process usually took multiple days.
Britta: To address these hurdles, we decided to move away from a pure spreadsheet-based approach toward a more systematic, data-driven, and visual method for analyzing our SARs for compound optimization studies. It was also a perfect fit for our #CropKey company-wide strategy.
Our goal was to develop a solution that was not only scientifically robust but also intuitive and visually engaging, enabling every member of a compound optimization project to conduct multi-objective SAR analysis on their own and collaborate easily.
“SAR Slides” project needs - Quickly generate, analyze and update SAR reports
Aurélie: We faced a similar challenge, as we were not able to generate SAR overviews efficiently. The process was very time-consuming and particularly challenging for newcomers to get up to speed on advanced projects. In addition, we have an increasing number of optimization parameters to consider, and generating and updating the SAR on all parameters was complex. Therefore, we decided to digitalize the workflow to be more efficient and have easy access to up-to-date and accurate SAR trends at any time.
“For advanced projects, it often took several hours to generate a complete SAR report and ensure all data points were up to date and all compounds were considered.”
As Britta and Janosch initiated the “MMPs” project with Discngine to develop a digital tool for multi-parameter SAR analysis, three years ago, we saw an opportunity to build upon it. Our goal was to create an application that automatically generates high-quality SAR reports, leveraging the methodologies established during the initial developments of SAR analysis and further expanding them (for instance, with SAR Matrices and R-Group deconvolution approaches).
Towards a comprehensive tailored solution
Key criteria of scientific solutions for SAR analysis and reporting:
- Standardized application: Once centralized platform designed for interdisciplinary drug discovery teams
- Data-driven multi-objective SAR analysis: Facilitates SAR analysis while considering multiple compound parameters
- User-friendly interface: Simple and intuitive, requires minimal training
- Information-rich dynamic visualizations: Use-case based visualization capabilities
- Flexible IT integration: Compatible with existing IT environments
“We hadn’t found any commercial solution that would fit our needs, so we had to create it. Discngine proved to be an excellent partner in delivering quickly on this high scientific need.”
Situation on the market and key criteria
Britta: Before starting custom development, we explored the existing solutions on the market. However, none met our key requirement for multi-parameter optimization or provided sufficient flexibility to integrate with our existing research IT landscape. At the same time, we were embarking on a digital transformation journey and needed a strategic and reliable partner. Ultimately, we decided to develop a tailored solution and chose Discngine to collaborate on the journey.
Janosch: The goal was to find the right balance between enabling complex multi-dimensional analysis and having an application that is as easy to use as possible and highly visual. The marketed products we explored, as well as existing internal tools, generally fell at one extreme or the other. Additionally, these tools primarily focused on analysis while overlooking crucial aspects like dataset preparation and sharing capabilities.
David: We chose to build on the “MMPs” project work as a foundation for our own SAR Slides project since none of the available commercial solutions fully met our needs either. Some tools for SAR reporting appeared outdated, and none provided the ability to display a substantial amount of information in a single, comprehensive view.
Product development journey
Britta: At the start of the project, with Discngine we defined a clear vision: to enable lab scientists to independently perform standard chemoinformatic tasks for SAR data analysis. Firstly, we successfully validated - with pilot studies - methodologies forming the foundation of the application, including Matched Molecular Pairs and Series (MMPs, MMSs). Afterward, we focused on developing a more visual and user-friendly interface tailored to the needs of chemists.
Janosch: There were many back-and-forth discussions with Discngine, especially early on when we recognized the lack of flexibility in existing technology. We decided to build a web-based application from scratch to enable greater interaction and collaboration. Through an iterative process, we arrived at the current version, which has crucial added value.
Aurélie: For the SAR Slide project, only three months after the project onset and mapping our requirements, we successfully developed a minimum viable product (MVP) for SAR reports generation. Using a real dataset under a confidentiality agreement allowed us to quickly cross-check and validate the tool’s functionality and accuracy. Most importantly, it allowed us to scale effectively, get into details, and capture the interest of a broader user base.
“We decided, jointly with Discngine, to reach an agreement for Discngine to create a commercial product as we believe many scientists can benefit from it and to also benefit from all future iterations and product evolution, and make sure it remains sustainable.”
The SAR Slides application has been fully productized and is now available for commercial use!
Synergistic solution for SAR data analysis and reports generation
The two projects initiated one after another have been merged into one application called PULSAR (Pilot Utility Library for SAR exploration) that now comprises two modules:
The “MMPs” module for multi-objective SAR analysis based on matched molecular pairs and matched molecular series methodologies
The “SAR Slides” module, for automatic SAR reports generation and visualization, based on MMPs and R-Group deconvolution methodologies
Britta: The solution we developed with Discngine fulfills all our predefined criteria. With our “MMPs” module we succeeded in creating a comprehensive tool for systematic, data-driven SAR analysis that integrates multiple SAR parameters simultaneously. Based on the matched molecular pairs algorithm, the tool enables efficient management and analysis of large volumes of bioactivity data.
Janosch: As I mentioned before, we were looking to find the right balance between complex analysis and high-quality visualization. We have managed to find that “sweet spot”, bringing complex use cases and analysis in a user-friendly, visual environment. These are now central features of the developed application. Users can quickly create multi-dimensional SAR analysis and find answers to their questions in a matter of hours compared to days before.
“About 80% of all our use cases now take a tremendously shorter amount of time than before - from multiple days to just a couple of hours for multi-dimensional SAR analysis.”
Aurélie: We are excited to have a contemporary tool – SAR Slides – that enables quick analysis and visualization of SARs on any dimension and any disconnection of a molecule. Users can now easily analyze and navigate SAR data, with the option to export SARs as images and use them in PowerPoint slide deck. This way, it supports the discussion on the project advancements during meetings with minimal preparation time.
“I can confidently say that creating SAR reporting has become incredibly efficient. Preparing datasets, uploading them, and generating SAR reports now takes just 15 minutes, compared to hours before. And if the dataset is already available, only seconds.”
SAR Slides application: Chemists can easily import and update their datasets, establish and visualize automatically organized SAR reports and associated metadata, ensure consistency in font size, style and alignment as well as identify trends and insights within datasets. The image is based on MMPs identification methodologies. (click to zoom in)
Wish to know more about the SAR Slides application? Visit our product page to get started
Discngine impact on solution development
Dual skilled expertise – science and IT – to quickly address the highly scientific need
“Overall, the biggest advantage of working with Discngine is their exceptional balance of scientific and technical expertise. This minimizes the need for translation between the worlds of scientists and IT, saving valuable time and speeding up project progress.”
David: Discngine possesses a rare expertise combination – scientific and IT - that truly makes a difference in scientific informatics projects like ours. They understand complex scientific concepts and have the developer skills needed to deliver effective solutions.
Unlike other IT-driven companies, Discngine immediately grasped the scientific challenges we faced, our requirements, and our end goals, which made it much easier to make project decisions and speed up solution development. Additionally, their proactive approach, genuine interest in adding value, and willingness to suggest both technical and scientific enhancements greatly contributed to the success of the project.
Regarding security and confidentiality, it was impressive that we could work directly with our own real data and address real use cases. Discngine demonstrated a high level of professionalism in handling client data safely and showcased extensive knowledge of security best practices.”
Janosch: Working with Discngine brought a super fluent stream going from science to technical implementations. Whenever critical project decisions needed to be made, Discngine quickly provided ad-hoc, tangible examples with rapid implementation, which greatly supported discussions and final decisions. Communication was smooth and regular, with many back-and-forth exchanges, all of which helped accelerate project timelines.
Trusted digital transformation partner
“I can confidently say that we were going through this digital transformation journey together, and over time, Discngine became a trusted partner in this strategic move”
Britta and Janosch: From a technical point of view, Discngine was a perfect match for our digital transformation journey. Their extensive experience in the research informatics field - working with other compound discovery companies - enables them to compare similar issues and quickly understand problems. They can then easily determine if previously developed solutions can be applied again, which helps in pushing forward digital solutions development. In our journey, we started with on-premises developments and gradually transitioned to cloud-based web applications, where Discngine demonstrated exceptional adaptability and flexibility.
Excellent support
Britta: Regarding support, the Discngine team proved to be very reactive and efficient. Whenever an issue arises, their response is immediate, and they manage to find a solution quickly. Our very close relationship allows us to address questions and inquiries via direct channels and regular meetings. This open communication extends beyond us as internal project managers. New team members also have the freedom to reach out directly, which is highly appreciated and fosters a sense of real teamwork.
Real-world benefits of the solution on compound discovery projects
Fast decision-making
David: “Chemists can now quickly and intuitively assess SARs and analyze several parameters simultaneously to question correlations. It streamlines and accelerates decision-making by providing clarity and insights on critical gaps, overlooked compounds, and the potential of compound subclass for their prioritization — insights that might otherwise go unnoticed. “
Janosch: The current solution helps tremendously with time saving, but also easily discovering potential issues to solve. Sometimes, scientists do not have clear or precisely formulated questions at the beginning. With the usage of the application, we started to transition from open-ended exploration - where it was unclear where to look or focus - to a more guided and structured analysis. The solution allows you to find your issue in the first place, along with the right follow-up questions for your next decisions. It connects insights that could be previously overlooked and allows them to better define and focus on the right questions.
Innovation
Aurélie: The SAR slide module is highly intuitive with a short learning curve. Even colleagues who are often hesitant to use new technologies find it appropriate. It facilitates greater involvement in projects, with new members catching up quickly on prior work, even at advanced project stages. Overall, it contributes to more people power and innovative idea generation.
Multidisciplinary work and collaboration
Britta: Chemists can now use a standardized tool to independently analyze SAR data. Additionally, data scientists and other disciplines can also view the data from a chemist’s perspective to easily understand science and the compound’s SARs - which was previously difficult. This capability promotes independent analysis and interpretation, enhancing collaboration within interdisciplinary project teams.
Conclusion
Structure-activity relationship (SAR) analysis is a cornerstone of compound optimization and a routine task for every discovery chemist and fellow data scientists. To enhance this process, Discngine and Bayer Crop Science joined forces on challenging yet important projects: the initial “MMPs” project for efficient SAR analysis, followed by the “SAR Slides” project for streamlined SAR overview and reporting. The result is a comprehensive solution – the PULSAR application – including two flexible, easy-to-use modules with complementary functionalities for multiple SAR analysis and reporting use cases. PULSAR integrates smoothly into Bayer workflows to drive collaboration and ideation in drug discovery.
Due to the high success of this project, the SAR Slides module has been fully productized as a stand-alone tool with strong integration capabilities. It is now available for commercial use! SAR Slides allows chemists from pharmaceutical, agricultural, and cosmetic R&D teams to quickly generate, visualize, and update their SAR report enhancing new molecule discoveries.
“The agile approach from Discngine, combined with intense and trustful interactions between scientists from Bayer and Discngine, has successfully transformed what started as an idea to enable SAR analysis for chemists into a tangible product.”
From customer project to fully commercial product! Learn more about SAR Slides application for efficient SAR reporting and analysis.
Contact us