Metadata harmonization, analysis pipelines, and AI foundational models together form the backbone of modern target discovery. By standardizing heterogeneous omics datasets, harmonized metadata enables meaningful cross-study comparisons and deeper biological context. Robust analysis pipelines ensure consistent, reproducible insights at scale, reducing time spent on data preprocessing. Building on this foundation, AI models can uncover hidden patterns, prioritize targets, and generate novel hypotheses. Collectively, these capabilities accelerate the transition from raw data to actionable insights, enabling more confident and efficient identification and validation of therapeutic targets in complex disease landscapes. This talk highlights our ongoing efforts to integrate these capabilities to accelerate target discovery.

This talk introduces Cheiron’s AI-native operating system for drug development, designed to unify scientific, clinical, regulatory, and competitive workflows. Powered by a Lifesciences Knowledge Graph, it connects diverse data sources and enables workflow-native AI agents, accelerating decision-making and streamlining biopharma processes from insight generation to execution.

The genetic complexity of congenital adrenal hyperplasia (CAH) makes it difficult to call variants of the disease. We will explore how machine learning was utilized to make this process easier and to narrow down on diagnoses.

An expert panel discussion on the challenges in validation, integration and scalability of AI models will follow the above talks. This interactive session will unpack some of the key bottlenecks in deploying AI models for data curation, integrating scalable AI workflows and how GPT-based chatbots are influencing this dynamic space.