Realize the Promise of Multi-Omics for Precision Medicine with Genedata Profiler
September 1, 2024
Marie-Ange Kouassi
Over the past decades, incredible advances have been made using omics technologies. These have improved the understanding of diseases and human biology, uncovering novel and more promising avenues for developing precision medicines. However, preparing, combining and analyzing these disparate, heterogeneous datasets remains laborious and inefficient. To overcome the challenges of working with multi-omics data, biopharma and biotech companies urgently require digital tools that streamline the conversion of integrated omics into meaningful scientific insights and increase their impact on patient care.
The Promise of Multi-Omics in Drug Development
Omics technologies provide a remarkable opportunity to explore human biology and have already enhanced our understanding of cancer, neurodegenerative, and inflammatory diseases.1,2,3,4,5 Assessing changes occurring on and the interplay between diverse biological levels such as genomics, transcriptomics, proteomics, metabolomics, epigenomics, and lipidomics, even on a spatial level has enabled scientists to identify molecular mechanisms underpinning disease, uncover disease subtypes, better predict response to treatment, or decipher mechanisms of drug resistance.6 Powerful insights derived from multi-omics data are also valuable in predicting future complications or adverse treatment effects allowing earlier intervention and better patient safety. Ultimately, these applications optimize drug development as they drive focus on the most promising targets, and the matching of a treatment to the right indication and patient population, improving clinical trial success and healthcare outcomes.7
Yet, extracting valuable insights from multi-omics data is no trivial task. This is because of the wide variety of instruments used which generates vast amounts of complex heterogeneous data of varying quality that require standardization and sophisticated algorithms for data integration. An important aspect of multi-omics data preparation is harmonized metadata annotation for efficient data identification, retrieval and maintaining data integrity. To derive insights and impactful insights from multi-omics data, scientists also require user-friendly analysis tools which enable in-depth exploration through sophisticated analyses such as Machine Learning. Such approaches offer advanced capabilities in pattern recognition and predictive modeling facilitating comprehensive interpretation of multi-omics data and the identification of hidden relationships. Due to the lack of easy-to-use analytics tools, scientists often need to rely on the technical expertise of data scientists to convert complex multi-omics datasets into sound conclusions. However, data scientists themselves can face challenges while working with these datasets if they lack digital solutions with the performance, scalability, and flexibility required for computationally intensive data mining and analysis.
Using the Genedata Profiler platform, inter-disciplinary drug development teams can expedite insight generation from multi-omics data emerging from any R&D stage, source or omics technology. Alongside the platform, teams can benefit from Genedata’s decades’ worth of knowledge and experience gathered through working with leading biopharma and biotech organizations such as Gilead, Merck KGaA, MacroGenics and Genmab. By consulting Genedata’s bioinformatics experts, data scientists can optimize their data mining and analysis approaches while scientists can receive guidance on how to best interpret results.
This article highlights how the award-winning platform Genedata Profiler, enables data scientists and scientists to overcome the challenges faced during multi-omics data analysis, accelerating the discovery of valuable insights that advance precision medicine (Figure 1).
Figure 1. Genedata Profiler supports the entire multi-omics data lifecycle. The platform provides the scalability, performance, and the flexibility for data scientists to use preferred programming tools and community-standard pipelines to process and analyze omics data. Non-coding scientists can also consume analysis-ready omics data in user-friendly dashboards and point-and-click analytical tools. Together, these capabilities equip scientists and data scientists to collaboratively generate insights that drive precision medicine.
Equipping Data Scientists with Data and a Flexible, Performant Computing Environment
Self-Service Access to Multi-Source Omics Data
Preparing high volumes of multi-omics datasets for insight generation can be challenging. This is partially because these data emerging from diverse instruments, often exist in disparate sources with different organization and privacy requirements. As a result, their retrieval for analysis can be extremely complex.To address this challenge, Genedata Profiler seamlessly connects to diverse data sources allowing for centralization of suitable omics datasets. These datasets are then organized consistently upon ingestion/federation and annotated with appropriate metadata for continuous data findability. Also, the project-centric organisation of data in the platform simplifies collaboration. At Merck KGaA, the use of Genedata Profiler as a central omics platform brought about a cultural shift democratizing access to data throughout the organization for different types of users. It was highlighted that “users feel now that they can get access to data outside of their own work area much easier than before “resulting in seamless data exploration and increased efficiency of biomarker research.8
Flexibility
Without standardizing routine data mining tasks, data scientists can lose excessive amounts of time reproducing efforts to generate harmonized, analysis-ready datasets. Genedata Profiler enables data scientists to standardize repetitive omics data processing activities using out-of-the-box yet configurable workflows, gaining efficiency and reproducibility through automation.
Often, data scientists differ in their way of working using preferred programming languages or omics data mining tools for specific types of tasks. It is thus important that they have a connected ecosystem encompassing the required data systems and a choice between widely used data wrangling and advanced analysis tools that enable machine learning and AI without the need to build and maintain a complex IT infrastructure. Such a secure, validation-ready ecosystem streamlines the activities of data scientists and reduces organizational and operational costs. Customers of Genedata Profiler benefit from added flexibility leveraging powerful programming languages including R and Python for working with data throughout its life-cycle.
Performance and Scalability
When handling such large and complex omics datasets, data scientists can face latencies if the infrastructure they are using is not built to support this. To avoid delays while working, it’s important that biopharma and biotech organizations equip their data scientists with a purpose-built platform that optimizes performance and costs even as datasets grow, and analyses become more complex. Referred to by Genedata customers as a “cloud-native solution” with “data durability, compute elasticity, platform resilience, cost efficiency and continuous validation”, Genedata Profiler provides data scientists with the efficiency they need to accelerate data analyses. This is due to its state-of-the-art analytics environment equipped with embedded elastic computing, unlimited storage, and the ability to save computing resources by placing data in “cold” storage while maintaining discoverability. Collectively, these capabilities and others provide the scalability required for growing R&D projects.8
Collaboration
Data scientists often work in concert with other non-coding subject matter experts, regularly communicating insights generated from omics data and providing harmonized, analysis-ready data in user-friendly applications. To provide them with interactive tools for omics data exploration and hypothesis testing, data scientists need an environment that allows application development and delivery in an efficient and reproducible manner. This way, data scientists can leverage Genedata Profiler to develop, test, package and deliver data analysis tools to scientists while maintaining traceability through code versioning.
Empowering Scientists with Self-Service Access to Analysis-Ready Data and Intuitive Analytics
Simplified Multi-Omics Data Exploration
To successfully test hypotheses and generate new ones, scientists need to be able to independently access all relevant multi-omics data and explore them in suitable purpose-built analytical tools. For this, scientists’ analysis tools must be easy to use without a steep learning curve in bioinformatics or programming. Genedata Profiler enables scientists to generate insights from omics data by offering easy ways to visualize these datasets in fit-for-purpose dashboards. For example, translational scientists at MacroGenics use interactive analytical applications of Genedata Profiler to routinely visualize heterogeneous high-dimensional patient molecular data gathered from oncology clinical trials. The insights generated from these recurring analyses help inform decision-making throughout clinical development.
Scientists may also want to leverage advanced sophisticated approaches to interrogate omics data further to e.g., identify novel biomarkers. For this, they require an intuitive analytical solution that would be easy to use yet enable comprehensive interrogation of multi-omics datasets. For instance, by using the powerful inbuilt no-code analytical solution of Genedata Profiler, MacroGenics’ scientists could access a wide range of statistical tests and the latest algorithms for comparative analyses and AI-based approaches, for deep interrogation of their multi-omics datasets. This enabled them to identify phenotypic, cellular, and molecular parameters associated with drug response allowing patient stratification during clinical development.
Leveraging Public Omics Datasets
Where statistical power is limited, scientists may look to publicly available datasets to validate findings discovered with data generated in-house. This can be challenging as the data may be difficult to access, may not be the latest version or may require harmonization before analysis. Genedata Profiler simplifies scientists’ use of public data by providing support for ingesting these data types and connecting them to user-friendly analytics tools. With easy access to the latest public datasets in analytical applications, scientists at leading biopharma organizations can interactively compare their in-house findings with public data e.g. TCGA to explore indications that could be responsive to a treatment in development. With Genedata Profiler as a data science ecosystem, scientists can gain faster access to user-friendly applications built by data scientist colleagues. At Merck KGaA, scientists benefit from over 150 analysis applications all managed within the App Space of Genedata Profiler (X-omics).8 From these apps, non-coding data consumers only consume data they have authorization to access and analyze thanks to Genedata Profiler’s advanced data governance capabilities.
Genedata Profiler-Driven Multi-Omics Data Analysis Applications
The possible applications of multi-omics analysis are manifold so long as one is equipped with a suitable software solution that simplifies the capture, analysis and interpretation of these heterogeneous datasets. Genedata Profiler has already facilitated the understanding of multi-faceted, complex diseases to enable precision medicine approaches. For pediatric Crohn’s disease, the platform streamlined the standardization of transcriptomic, proteomic, and clinical data from a pediatric inception cohort. In addition, Genedata Profiler provided machine learning and advanced statistical ranking approaches to identify and validate an optimized set of biomarkers to predict disease course and response to therapy. Biomarkers identified using Genedata Profiler through combining gene expression with clinical data also formed the basis of a precision medicine tool to predict the response (or lack of response) to infliximab, the biological treatment for Rheumatoid arthritis.9 Moreover, Genedata Profiler led to the development of an unbiased machine learning model to detect pancreatic cancers in patient subgroups based on epigenetic DNA modifications.10 For assessing the suitability and safety of treatment for patients, Genedata Profiler was also used to obtain a mechanistic understanding of toxicological responses in specific target organs. For this, the platform was used to process raw omics data and build a bioinformatics pipeline to analyze transcriptomic, proteomic, genomic and metabolomic data generated from in vitro experiments consisting of therapeutic-exposed hepatic and cardiac 3D microtissues.11
Conclusion
With drug development costs increasing and the rate of successful treatments delivered to market decreasing, it’s clear our understanding of disease pathobiology and associated molecular drug targets needs improvement. Multi-omics analysis is the key to gaining deeper insights into how complex diseases develop and progress over time. Rather than investigating changes in individual omics in isolation, multi-omics analysis allows us to examine the interplay between different molecular counterparts, providing an overall more complete understanding and accurate deductions. Guiding drug development with multi-omics insights allows the identification of more precise therapeutic targets and strategic patient selection for clinical trials. This should allow drug development organizations to proceed with precision and develop promising drug candidates faster, with a higher probability of clinical success. To make this a reality, leveraging the right digital platform that empowers different stakeholders is key. This would help biopharma organizations unlock untapped omics data potential through groundbreaking analytical approaches, accelerating precision medicine.
Contact us to find out how you can leverage Genedata Profiler today for multi-omics data analysis.
References
- Irmisch A, Bonillia X, Chevrier S. The Tumour Profiler Study: integrated, multi-omic, functional tumour profiling for clinical decision support. Cancer Cell, March, 2021, 39(3): 288-293.
- Nativio R, Lan Y, Donahue G. An integrated multi-omics approach identifies epigenetic alterations associated with Alzheimer’s disease. Nature Genetics, October, 52(10):1-24-1035.
- Lee A, Kim C, Park S. et.al Characterization of altered mechanisms in Parkinson’s disease through cell type-resolved multiomics analyses. Science Advances, 2023: 9(15)
- Wang K, Khoramjoo M, Srinivasan K. et al. Sequential multi-omics analysis identifies clinical phenotypes and predictive biomarkers for long covid. Cell Reports Medicine, 2023 Nov 21: 4(11)
- Chen J, Li S, Zhu J, et al. Multi-omics profiling reveals potential alterations in rheumatoid arthritis with different disease activity levels. Arthritis Research & Therapeutics. 2023, 3:25(1):74
- Athieniti E, Spyrou G. A guide to multi-omics data collection and integration for translational medicine. Computational and Structural Biotechnology Journal. 21(2023), 134-149.
- Kraljevic S, Stambrook P.J, Pavelic K. Accelerating drug discovery. EMBO reports. 2004, 5(9): 837-842
- Pharma-Genedata Collaboration Builds ROI-Rich Tool for Omics Data Exploration, Bio-IT World News, May 31, 2022.
- Genedata has successfully collaborated with Egis Pharmaceuticals PLC to develop a precision medicine tool for Rheumatoid Arthritis. PR Web, Oct, 2021.
- Hartwig C, Müller J, Klett H. et al. Discrimination of pancreato-biliary cancer and pancreatitis patients by non-invasive liquid biopsy. Molecular Cancer, 23, 28 (2024).
- Verheijen M, Sarkans U, Wolski W. et al. Multi-omics HeCaToS dataset of repeated dose toxicity for cardiotoxic & hepatotoxic compounds. Sci Data 9, 699 (2022).