In this talk, two case studies will demonstrate how Pfizer data scientists use the power of automation to analyze and understand the vast quantities of data that are generated on a daily basis. Discussion topics will include why data automation is so useful, data automation strategy, and the impact of two data automation projects that were completed at Pfizer.

This presentation will highlight Janssen BioTherapeutics development digital maturity journey. Beginning with defining a holistic program partnering with IT to build a scalable, supportable infrastructure and deliver useful tools/apps/data access to project teams. Approaches to enabling the data to flow with context will be shared with examples of key successes and opportunities ranging from structured data capture to applied data science and e-reporting.

As a central interface in Roche pRED’s vision of the lab of the future, AutoLab enables seamless and automated data flow between instruments and various data sources. AutoLab increases research efficiency by providing easy-to-use digital workflows, guiding the scientists in their daily work and automating several time-consuming tasks. The digital representation of lab and data workflows supports FAIRification of data collected along the entire value chain.

• At what stages and areas of analytical development do you see big opportunities for digital applications?
• Success stories, major challenges and your solutions
• Types of modeling applied for analytical development and the value they bring
• Different needs for different modalities

• Current state assessments: what are we doing and how - and what aren’t we doing
• Data capture and standards: pain points and opportunities (ELN, systems, instruments & context)
• Setting digital goals: what has worked and at what levels of detail
• User requirements: best approaches for science and engineering
• Data flow end user experiences: good or bad and lessons learned

The success or failure of projects centered around digital transformation can be determined overall by an organization’s vision, alignment, resources, and capabilities. Focusing specifically on digital integration in busy research laboratories, we’ll explore how these themes manifest and interact to shape the outcomes and impact of projects.

Data science and digital applications are being more widely used nowadays to accelerate analytical development. In this talk, a few case studies will be shared to show how CMC statisticians, data scientists, analytical scientists, and IT partners work together to implement solutions that help gain deeper insights, quantify risks, enable quality-by-design and data-driven decisions, and improve efficiency.

Model-based prediction of biologics developability will increase speed to clinic. Previous program data can be used to train models but requires data quality control and compensation for biased sampling of sequence space. In my talk, I will describe how we incorporated historical data using data quality control protocols and used sequence similarity clustering to improve prediction of critical quality attributes of monoclonal antibodies.

As Process Analytical Technology (PAT) matures and becomes more robust, it offers more incentive for companies to integrate PAT into real-time processing. Additionally, when PAT is implemented, it opens up additional opportunity for companies to explore higher level of control. In this presentation, we will share a case study of how PAT can be integrated with Model Predictive Control to improve bioprocess efficiency. 

In silico assessment of antibody developability has the potential to speed up antibody development, especially lead optimization. However, advances in computational tools such as machine learning are often limited by the lack of suitable training data sets of high quality. Here, I present improvements of common antibody developability assays, e.g., AC-SINS, with the aim of enabling optimal data for modelling. Important parameters like dynamic range, calibration, and data processing are discussed.

This talk will address the benefits of utilizing smart data models (ontologies) in transitioning to Biopharma 4.0. First, a comparison between ontologies and traditional data standards will be given. Next, success stories of utilizing ontologies will be outlined, along with the most prominent ontologies in the field. Finally, the application of ontologies to two critical use cases will be demonstrated: cross-domain data integration and improvements in data-driven and hybrid modeling.

With Pharma 4.0, the digital revolution, analytical scientists have come to expect access to their scientific information in much the same way as they do with their personal devices. More than just accessing data, scientist want to use their information to gain insights, build applications, and make their work empowered by the very information they generate. To meet this demand pharma companies will need to overcome challenges in data access, making their information FAIR, and developing the skills necessary to accelerate their science in the digital age.

Throughout years of cross-functional collaboration, research and development of each pharmaceutical candidate accumulates tremendous amounts of data. However, due to the segregation and heterogeneity of these data, knowledge extraction has been tedious and limited while dossier authoring becomes an excruciating exercise. Enterprise data lake and ontology enabled data curation and semantic transformation enables structured content and data management that democratizes scientific data, enabling deep data analytics and automated dossier authoring.