PMx Data Programming SIG | AI-Driven Non-Compartmental Analysis Dataset Preparation

Recorded On: 09/18/2026

This webinar will introduce an AI-enabled approach to preparing analysis-ready Non-Compartmental Analysis (NCA) datasets for pharmacokinetic parameter generation.

NCA estimates pharmacokinetic exposure parameters from observed concentration–time data without assuming a specific compartmental model.

The NCA input dataset is the analysis-ready deliverable that integrates PK concentrations, sample collection and dosing information, demographics, metadata, and study-specific rules.

It serves as the foundation for PK parameter generation, CDISC dataset creation, and subsequent TFL generation to support the clinical study report.

To accelerate and standardize NCA dataset preparation, the team has implemented AI-driven automation that addresses repetitive, study-specific programming activities.

The business need stems from the consistent interpretation of source data, specifications, and data standardization workflows across studies.

By combining data science capabilities with PK subject matter expertise, the team is developing a scalable application that automates routine dataset generation while improving speed, consistency, and transparency.

A key design principle is that AI generates a transparent, study-specific configuration rather than executable code.

The dataset-building pipeline remains fixed and validated, while AI derives the settings that drive the pipeline from each study’s specifications and source data.

Each AI-derived decision is auditable and surfaced to the PK SME for review, preserving human oversight for interpretation while reducing repetitive programming burden.

Weidong Chen

Bristol Myers Squibb

Weidong Chen is an Associate Director in Data Science within Translational Medicine & Clinical Pharmacology at Bristol Myers Squibb. Since joining BMS in 2015, he has provided data science and programming leadership for pharmacometric analyses supporting early drug development, late-stage submissions, regulatory filings, and internal decision-making across Immuno-Oncology, Immunology, and Cardiovascular studies. Weidong brings deep expertise in SAS programming and clinical trial data standardization. Earlier in his career, he held statistical and SAS programming roles at Frontage Labs and Allergan, supporting Phase I/II, Phase IV, and bioequivalence clinical trials across multiple therapeutic areas. He holds a PhD in Biochemistry and Molecular Biology from Stony Brook University.

Amir Molavi

Bristol Myers Squibb

Amir Molavi is a Senior Software Engineer and Computational Scientist on the Data Science team at Bristol Myers Squibb in Cambridge, MA, where he supports AI, generative AI, and machine learning efforts within TMCP. His work focuses on bringing modern AI capabilities into scientific and data-driven workflows. Amir has a strong foundation in computer programming, physics, and mathematics, with deep expertise in Python, numerical modeling, and machine learning. Before joining BMS, he spent four years at Applied BioMath, where he redesigned the backend and Python client API powering model-informed drug discovery simulations. He previously conducted computational research as a postdoctoral associate at Northeastern University. Amir holds a PhD in Structural Engineering and an MSc in Applied Mathematics from the University of Akron, along with certifications in Machine Learning and Deep Learning from the University of Texas at Austin.