In an era where biopharma is inundated with volumes of multiomics data, the challenge extends beyond data collection to robust analysis, scalability, and effective integration. This presentation offers an overview of software solutions that merge cloud computing, data engineering, and custom workflows. Rooted in real-world challenges, we’ll explore the trajectory from raw data to actionable insights, ensuring data security, shareability, and enterprise adaptability along with insights gained from customer-led software development.

This talk explores the potential of data science through the strategic integration of Large Language Models (LLMs) and software skills. Explore how this symbiotic relationship accelerates workflows, facilitating efficient data analysis and interpretation. Join us to grasp the transformative impact of synergizing LLMs and software expertise, offering a gateway to enhanced productivity and innovation in the dynamic realm of data science.

Federated Learning (FL) enables creation of more robust models without the exposure of local datasets. However, FL does not guarantee the privacy of data, because the information extracted and utilized to infer the private local data is used for training. We developed Advanced Privacy Preserving Federated Learning framework (APPFL), with advances in differential privacy, to enable Privacy-Preserving Federated Learning (PPFL). We enabled training of AI models in a distributed setting across multiple institutions, where sensitive data are located, with the ability to scale to help create robust, trust-worthy AI models in biomedicine applications where data privacy is essential.

A collaboratively architected and implemented flexible platform created by Nebulaworks cloud engineers and a pharma client established an environment to deliver bioinformatics services that ensures data analysis results and metadata are preserved. A code-first, easily deployable containerized technology design provides data services, analysis services, and AI/ML-readiness. The architecture enables bioinformatics staff to focus on data integration and analysis that support asset evaluation from discovery through to clinical development.

The field of Data Science is characterized by its rapid evolution, marked by daily developments that hold the potential to significantly enhance the expeditious conduct of clinical trials. In this context, the implementation of a dynamic data science platform built upon a polycloud architecture becomes indispensable, as it empowers data scientists to harness the cutting-edge capabilities offered by premier cloud service providers. Furthermore, the integration of advanced data visualization capabilities into this platform serves to foster seamless collaboration among researchers. Learn valuable insights from our experience in implementing these platforms within a large pharmaceutical company, where they are utilized by hundreds of individuals. Our presentation will shed light on how we've successfully deployed this dynamic platform to cater to our users' evolving requirements amidst a rapidly changing technological landscape. We will demonstrate our approach to fulfilling diverse experimentation use cases leveraging Multiverse/Polycloud Data Science Platform

Biological sciences has a reproducibility problem related to the bottleneck created by the over simplification of biological modeling. This talk discuss methods developed to improve the analysis of biological interactions so that more nuanced understandings of physiological interactions can be  produced. Learn about problems that have been caused by incomplete understanding of drug interactions specifically related to Alzheimer's disease drug development. This will be illustrated by describing enzymatic mechanism of action studies and how these studies can benefit from automation of the analytical process.

The massive expansion of biological data collections in the last 20 years alongside dramatic developments in artificial intelligence has created an unprecedented opportunity for using data to integrate human and mammalian model research, unlock biological insights, and empower the biomedical community worldwide. To meet this challenge, Data Science at The Jackson Laboratory is leveraging our unique expertise in mammalian genetics to unlock and integrate robust, but currently unconnected or unavailable, life sciences datasets for data-driven discovery. This talk will discuss the data science opportunities for transforming how we incorporate model data into biomedical discovery.