With each year, the number of curated and community sources of publicly available data and repositories increases. While this is great news for the whole scientific community, it can be quite overwhelming for non-expert users. Here, we present an approach using semantic technologies and heavily rooted in the FAIR principles to make data more easily available to scientists, with a special focus on enriching internal data with publicly available data in the biologic space.

When we set out to build a data strategy it was a loose set of ideas together with a big vision. Three pillars of execution have helped to turn it into action: 1) The scientific question as the main framework for execution, 2) building a team with diverse skillset that fully understands the data, 3) integrating data governance to shape data citizenship and a novel way of data management.

Join to hear firsthand from Sangeet Khullar, Director of Data Architecture at Daiichi Sankyo, as he walks through his experience on Data Quality, RDM, and MDM initiatives in Pharma, including:

• Reference data ( industry-standard dictionaries for data analysis and publications)
• Master Data – Consistency of Protocol Numbers 
• Cost reduction through simplifying data processes
• Regulatory Mandates ( IDMP etc.)

The session will detail how Daiichi Sankyo uses the Ataccama platform to control the quality of their data and maintain strict, holistic data governance—thereby supporting and ensuring the success of drug development processes.

In March 2021, Dotmatics, provider of enterprise scientific informatics software, joined forces with Insightful Science, holder of best-of-breed desktop software such as Geneious, SnapGene, and GraphPad Prism. This combination gives life sciences research organizations the compelling benefits of centralized data management provided by Dotmatics, while still allowing end-users to innovate in the tools they are familiar and productive with. This marks a major step forward for the industry.

Scientific organizations continue to strive for digital transformation and connectivity across their ecosystem. The need for better automation is apparent, including physical automation of the lab and the intelligent flow of information across the disparate digital tools that support their scientific and business processes. Automation allows for reduced downtime, increased stability, and optimized efficiency, providing the time and space for labs to focus on innovations and improvements crucial to their success. Through digital solutions, lab automation, and artificial intelligence, automation platforms are critical to the digital transformation journey.

There has been an explosion in the availability of data from various sources and formats and in the use of data from simple analytics to complex AI. And virtualization of all operations has accelerated our engagement with data. Governance is key to accessibility and quality data. We hope to uncover lessons in data governance from a selection of industries that deal with diverse data and problem spaces.

Transparent research practices and the sharing of research data, protocols, and code accelerate scientific advancements and solve real world problems in healthcare, the environment, and society. Yet, broad, open, research sharing has presented trust issues, particularly with limited checks on scientific integrity. This workshop will engage attendees on common trustbusters in open science and dive deep into the taxonomy of trust - providing research integrity recommendations and best practices.

With the adoption of CDISC data standards by the USFDA and PMDA for new, investigational, and abbreviated new drug applications, and certain biologics license applications filings, most biopharmaceutical companies are converting data collected in clinical trials to comply with these standards. Successful methods for accomplishing and managing this process using a singular data model across biopharmaceutical research will be discussed and will highlight the benefits to research.

Biobanks offer a rich repository of information to study for greater understanding of disease mechanisms and the creation of more targeted treatments. While the growth of such data sets has been impressive to those wanting to analyze it, it can also be daunting in scale. GPUs dramatically accelerate the analysis of multi-omic data, from secondary and tertiary analysis of sequence data, to deep phenotyping image analysis, to natural language processing of records. Acceleration makes at scale analysis of data not only tractable, but in some cases interactive.

The rising availability of multi-omics data for disease research represents an incredible opportunity for clinical research.  Automated harmonization of ICD-10 diagnostic codes and disease identifiers in omics data allows research data and clinical data to be integrated at scale.  This automation is the key to providing high quality data for ML applications in multi-omics clinical research.  We demonstrate automated harmonization and integration and its potential to unlock new research opportunities.

The development and deployment of AI and ML models in healthcare includes challenges that extend well beyond data availability and technology stacks. Data governance, regulatory compliance, stakeholder priorities, financial incentives, sustainability, and translational and clinical impact all affect if and how AI and ML research can benefit a learning healthcare system. In this talk, I will review some of the general challenges to translation of AI and ML into clinical practice and provide specific context based on experiences at our institution.