Agile workflows require transparency and easy-to-do sharing of ongoing activities across systems, processes, and departments. In this presentation I will talk about the learnings in a multi-year project to establish digital self-service processes of an analytics-heavy development function. Transparency for digital artefacts scattered across many islands-of-knowledge was achieved through an unusual combination of data insight services: data catalogue, knowledge graph, and Kanban boards.

While data platforms, data lakes, data swamps, and any other container for data you can think of seem to be the rage, we need to understand how to move upward on the ladder from swimming with data to powering a flight that is impactful on wings of knowledge and insight. In this presentation we’ll talk about mechanisms and techniques to learn, implement, and capture the process of climbing the ladder of inference in a knowledge-driven way to impact the drug discovery process.  We’ll look at examples of how capturing and codifying the process of knowledge extraction at scale allows the development of applications that yield rapid insights and how these tools free our scientists to leverage those insights to impact their work.

In an ever-changing IT landscape, it is a challenge to manage the core business knowledge. As systems change, data cannot be migrated/accessed easily, especially when a mix of validated/non-validated/on-premise and cloud systems are involved. How to provide a document management process that is flexible enough to follow the trend of moving data to cloud platforms, reflect organizational changes, and still retain the knowledge from proprietary systems? Roche Technical Development managed to find a structured home for millions of product-relevant documents, no matter where they are stored.

The life sciences industry is undergoing a digital transformation. Or, is it? Much has been written about the promise of data and analytics and the dramatic impact that they would have on the drug discovery process if only we learned to embrace the changes required to effectively realize the potential. In this talk, we will explore the evolution of data, analytics, and some potential future scenarios for the digital transformation to come.

For years, healthcare and life science have been making good use of their structured data. However, the larger part of enterprise data, nearly 80 percent, is unstructured and has been much less accessible, locked in silos across the data ecosystem. Embedding natural language processing (NLP) is the key to unlocking a step change in insight generation, decision making and innovation. In this session, we will discuss how flexible deployment of NLP capabilities can enable organizations to transform processes, increase efficiencies and realize meaningful value. A focus on use cases in drug safety will be presented, showcasing recent developments for cloud-first strategies and end user review.

FAIR is great for data...but what can be done to make text - fundamental research, scientific papers, lab notebooks, clinical results, and everything else that is text - FAIR? Learn how FAIR data principles also apply to unstructured content, using NLP and ML to turn unstructured text into structured information, and how search brings focus to the overwhelming sea of content. That's the key to innovation - connecting the right information with employees to accelerate innovation throughout the enterprise, including scientific discovery, development, clinical trials, or regulatory affairs. With an overview of the principles and several case studies, see how Sinequa is helping propel the life sciences into a new renaissance of information utility. 

PrecisionFDA is a secure, cloud-based, high-performance platform implemented by the FDA for collaboration in the areas of biological, clinical and chemical informatics in order to advance precision medicine. It has expanded to assist the broader community of data scientists for analysis of clinical data and to host data-driven competitions that push the limits of tools and algorithms for large-scale data processing. PrecisionFDA has hosted 31 challenges and app-a-thons on topics including AI/ML development to identify COVID-19 risk factors and predict brain cancer patient outcomes and bioinformatics tool development and benchmarking for genetic variant calling and detection of microbial pathogens. The latest application of this platform provides analysis of chemical data sets, as well as benchmarking computational approaches to cheminformatics, drug discovery, and toxicology.

Rare disease patients suffer too often from long diagnostic delays and misidentified diseases. This creates a significant burden, not just for patients, but for healthcare systems. We present in this talk examples of instances where we have collaborated with researchers and hospital systems to develop novel approaches for rare disease patient identification using tools like genomics, machine learning, and NLP.

Human genetics is maturing w/ robust insights from both monogenic rare variant genetics and population genetics converging with the increased impact of population scale WES and WGS data sets. This talk will give an overview of drug discovery approaches and insights from large consortia and how genetic data can be integrated with transcriptomic, proteomic, epigenetic and imaging data to gain detailed mechanistic insights from genetic hits, as well as give some examples of how AI and Machine learning is used in the process.   

We discuss the concept of deep digital transformation in pharma via data- and AI-enablement. Today, we are increasingly capable of integrating multiple modalities of data from clinical to omics and real-word evidence for drug discovery and development. How do we align our AI and digital strategies to produce value with end-to-end AI at scale. We determine ROI of AI in pharma with a real-life R&D use case in digital pathology. 

There is an increasing sense of urgency to implement transformative change within Pharma R&D to improve Real-World outcomes for all patients. The availability of RWD sources continues to grow, but to date, these sources still lack critical information on social determinants of health and behavioral factors at scale that can help improve our understanding of how diversity impacts health care. We will share our experiences in solving for this data gap.

Between discovery and clinical research lies translational, which advances science from laboratories to patients. We generate and consume heterogeneous pan-omics preclinical and clinical data from a multitude of internal and external sources, systems, and applications. To reach our goals requires wholistic data interoperability. We have adopted a data-centric approach that enables frictionless reuse of data across our systems, platforms, and applications. Key steps and lessons learned will be described.

We generated molecular profiling data for cancer exceptional responders from previous clinical trials, then conducted various computational analyses and identified key genetic/genomic features that could help to understand why some cancer patients responded very well for some drugs.

The literature and clinical trial landscape is growing in complexity and with increasing speed. Applying and optimizing artificial intelligence to tame information overload can lead to speedier, more personalized, and advanced searches compared with traditional methods. This session will discuss practical insights into novel platforms as well as provide perspective on the importance of communications and how to bridge discussions amongst healthcare stakeholders.