AI In Pharma

In this episode of The Deep Dive, we explore the groundbreaking application of quantitative systems pharmacology (QSP) to one of medicine’s most complex challenges: inflammatory bowel disease (IBD). Guided by research led by Katherine V. Rogers and colleagues, we unpack how advanced computational models are helping scientists understand the tangled web of immune pathways in Crohn’s disease and ulcerative colitis.You’ll learn how researchers built a dynamic, mechanistic model that simulates the human immune system in the gut capturing key players like T-cells, cytokines, and neutrophils and used it to mirror real-world patient responses to treatments like infliximab and ustekinumab. We explore how this virtual patient population can help identify likely drug responders, test combinations, and refine future clinical trials; all without stepping into a lab.This episode isn’t just about math and molecules. It’s about a new way of thinking in medicine, and how computational tools are shaping the future of drug development and precision care.

In this episode, we delve into how a multi-scale Quantitative Systems Pharmacology model is transforming our understanding of metastatic melanoma. You’ll hear a detailed discussion on integrating lesion-, patient-, and population-level dynamics to uncover hidden progression drivers, dissect mechanism-specific effects of checkpoint inhibitors, and guide the design of next-generation combination therapies. A must-listen for anyone interested in the quantitative modelling revolution in immuno-oncology.Here is the link to the paper : https://ascpt.onlinelibrary.wiley.com/doi/epdf/10.1002/psp4.12637

Deep dive into the origins, rationale, and practical implementation of quantitative systems pharmacology (QSP), structured around the six-stage workflow first articulated by Gadkar et al. (2016). Key highlights include:Introduction to QSP & MotivationA concise overview of QSP’s role at the interface of pharmacology, systems biology, and engineering, emphasizing the need for standardized workflows to improve reproducibility and stakeholder communication.Stage 1: Project Needs & GoalsDiscussion of how to engage collaborators, define decision-making timelines, and scope project questions so that modeling efforts align with real drug-development milestones.Stage 2: Reviewing the BiologyGuidance on literature mining, expert interviews, data aggregation, and visual diagramming to delineate the biological scope and identify knowledge gaps before building any equations.Stage 3: Model Structure DevelopmentExamination of approaches—supervised vs. unsupervised, logic-based vs. differential equations—to translate biological diagrams into mathematical topologies, with examples of pathway and multiscale models.Stage 4: Calibration of Reference SubjectsInsights on sensitivity and dynamical analyses, parameter estimation strategies, and the use of a small set of “reference virtual subjects” to ensure the model can recapitulate core behaviors.Stage 5: Exploring Variability & UncertaintyDescription of generating alternate parameter sets (virtual subjects), assembling virtual cohorts, and weighting them into virtual populations to capture heterogeneity and test predictive robustness.Stage 6: Experimental & Clinical Design SupportHow model outputs inform optimal experiment design, biomarker selection, and clinical trial simulations, and how new data feed back into iterative refinement.Concluding ThoughtsEmphasis on the cyclical, collaborative nature of the workflow and the value of “wrong” predictions in generating new biological hypotheses.

The podcast provides a deep dive into how Quantitative Systems Pharmacology (QSP) is transforming drug development by offering a systems-level view of disease biology and drug action. Key points discussed include:Understanding Mechanism of Action (MoA): QSP models enable detailed exploration of how drugs interact dynamically with biological systems, moving beyond static target identification.Simulating Disease Progression: Building "virtual disease models" allows researchers to understand and predict the natural course of a disease and simulate drug interventions.Virtual Patient Populations: QSP enables the creation of diverse virtual patients, capturing variability in genetics, physiology, and disease, crucial for predicting heterogeneous drug responses.Dose Optimization: QSP helps optimize dosing strategies rather than simply identifying the maximum tolerated dose, aligning with modern regulatory expectations like the FDA’s Project Optimus.Bridging Preclinical to Clinical: QSP supports translational modeling, helping bridge the gap between animal studies and human clinical outcomes by modeling key biomarkers and pathways.Model Qualification: Ensuring models are “fit-for-purpose” is critical—through rigorous validation, transparent assumptions, and biological plausibility checks.Decision-Making Support: Well-qualified QSP models inform early go/no-go decisions, optimize trial designs, and reduce risk in drug development.The episode concludes by emphasizing the importance of interdisciplinary collaboration (biology, modeling, pharmacology, mathematics) to fully realize QSP’s transformative potential.

In Supporting Drug Development for Rare Diseases using QSP, we explore how Quantitative Systems Pharmacology is transforming every stage of therapeutic discovery and development for conditions that lack traditional commercial focus. Each episode dives into the power of multi‑scale mechanistic modeling to connect molecular interactions with cellular, tissue, and whole‑body dynamics, especially when patient data are scarce. You’ll learn how virtual patient cohorts and in silico trials can predict efficacy and safety across diverse subpopulations, reducing reliance on large clinical studies. We unpack strategies for identifying novel targets and repurposing existing compounds, optimizing dosing regimens and trial endpoints, and ensuring regulatory readiness through standardized model qualification. Along the way, we spotlight cutting‑edge AI advances—rapid literature mining, automated data extraction, and intelligent trial‑design workflows—that turbocharge model construction, hypothesis generation, and decision‑making. Whether you’re a systems pharmacologist, drug developer, or curious innovator, tune in to discover how QSP and AI are enabling faster, more efficient, patient‑centric therapies for rare diseases.

The field of QSP is rapidly evolving, especially with its increasing integration with Artificial Intelligence (AI), the QSP modeling, at its core, aims to describe biological relationships mathematically, providing a mechanistic understanding of drug actions across multiple scales. Here are some key themes emerging:•Enhanced Drug Development: QSP plays a crucial role in various stages of drug development, from understanding disease mechanisms to predicting clinical outcomes and optimizing dosing regimens. Its application is being seen from target identification to clinical trials and regulatory submissions.•The Power of AI/ML: The synergy between QSP and AI/ML is unlocking new possibilities. AI/ML can assist in knowledge discovery from vast amounts of literature, aid in model building and parameterization, enhance the generation of virtual patient populations, and even contribute to hypothesis generation. This integration can accelerate the modeling life cycle.•Regulatory Acceptance: Regulatory bodies like the FDA are increasingly recognizing the value of QSP in drug development and review processes. There's a growing emphasis on "fit-for-purpose" models and the establishment of best practices for their development and qualification•Best Practices and Collaboration: The community is actively working on defining best practices to maximize the use and reuse of QSP models, emphasizing transparency, documentation, and interdisciplinary collaboration. Effective communication between modelers and stakeholders is crucial.•Applications Across Diseases: QSP modeling is being applied to a wide range of therapeutic areas, including neuropsychiatric disorders, immune-oncology, and rare diseases. The convergence of QSP with AI holds immense potential to improve efficiency, reduce attrition rates in drug development, and enhance our understanding of drug mechanisms and patient variability. It's an exciting time to be in this interdisciplinary field!