TL;DR
- Ionis partnered with Intelligencia AI to modernize PTRS benchmarking for portfolio decision-making.
- Replaced static, literature-based benchmarks with dynamic, asset-specific comparator cohorts.
- Benchmarks now incorporate indication, modality, genetic target, and development path for higher relevance.
- Introduced time as an explicit variable, enabling more current and regulatory-relevant comparisons.
- Added transparent PTRS drivers to explain underlying risk factors beyond headline probabilities.
- Improved clarity, consistency, and confidence in portfolio strategy discussions.
- Shifted PTRS from an external input to an internal, continuously evolving decision-making capability.
The Challenge: Static Benchmarks in a Dynamic Scientific Landscape
At Ionis, portfolio decisions are grounded in rigorous science and a deep understanding of clinical and regulatory risk. Central to this decision-making is the probability of technical and regulatory success, or PTRS, which serves as a critical input into how programs are prioritized, advanced, and resourced across the portfolio.
However, prior to working with Intelligencia AI, the portfolio planning team faced growing limitations in the PTRS benchmarks available to them. Existing sources were largely static and generalized, often drawn from published literature, review articles, or broad vendor datasets. These benchmarks were typically segmented by high-level categories such as therapeutic area or modality and did not reflect the specific characteristics of Ionis programs.
As Ionis increasingly advanced genetically targeted therapies, particularly in complex areas such as neurology, the shortcomings of generic benchmarks became more apparent. Many programs followed non-standard development paths, including direct transitions from early-stage trials to Phase 3, adaptive designs, and biomarker-driven strategies. These nuances were rarely captured in traditional benchmark frameworks, leaving gaps between how Ionis actually executed development and how success probabilities were being estimated.
This created two interconnected challenges. First, PTRS starting points lacked the level of specificity and currency needed to inspire confidence in high-stakes discussions. Second, the team had limited ability to respond quickly and credibly to targeted questions from leadership about how benchmarks should shift based on disease, modality, development path, or recent changes in the clinical landscape.
Ionis needed a more tailored, dynamic, and transparent approach to benchmarking and PTRS estimation, one that could evolve alongside both the portfolio and the broader scientific and regulatory environment.
“It ended up with this range of uncertainty of other benchmarks that was essentially useless for our comparative purposes.”
— Ionis Portfolio Planning Team
The Solution: Dynamic, Asset-Relevant Benchmarking and Explainable PTRS
Ionis engaged Intelligencia AI in the fall of 2024 as part of a broader effort to strengthen and modernize its PTRS process. Over the following year, the portfolio planning team integrated two core offerings into their workflow: Dynamic Benchmarks and the Portfolio Optimizer with PTRS estimates and drivers.
Building Benchmarks That Reflect Real Development Paths
Dynamic Benchmarks enabled the team to move beyond static, one-size-fits-all success rates and construct comparator cohorts that more closely resembled Ionis assets. Rather than relying on a single benchmark for a broad category, the team could apply multiple layers of filtering, including Indication, modality, genetic target, and development path.
This flexibility was particularly valuable given Ionis’s focus on genetically targeted therapeutics. Benchmarks could be constructed not only around similar modalities such as antisense oligonucleotides or siRNA, but also around shared gene targets across different therapeutic approaches. Development paths that deviated from traditional phase sequencing could be explicitly accounted for, ensuring that programs were compared against relevant analogs rather than forced into ill-fitting templates.
The ability to iterate on benchmark construction in real time fundamentally changed how the team prepared for and participated in strategic discussions. Filters could be adjusted, expanded, or relaxed to test the stability of results and avoid overly narrow or overly generic comparator sets. This created a more systematic and consistent approach to benchmarking across programs and reduced the risk of unintentional bias introduced by ad hoc cohort selection.
“I legitimately just pulled up the benchmarking tool on my computer and got the numbers within two minutes.”
— Ionis Portfolio Planning Team
Making Time an Explicit Dimension of Risk
Temporal relevance emerged as a critical factor, particularly in fast-evolving therapeutic areas. Static benchmarks often blended data from vastly different regulatory eras and technological contexts, obscuring meaningful signals about current probabilities of success.
With Dynamic Benchmarks, time became an explicit and adjustable dimension of analysis. The team could restrict cohorts to recent development periods, such as the past seven to ten years, aligning benchmarks with contemporary regulatory expectations and standards of care. Sensitivity analyses across different time windows allowed the team to understand how much results depended on historical versus recent data and to avoid letting outdated failures disproportionately influence present-day assessments.
Integrating External PTRS Estimates and Drivers
In parallel with benchmarking, Ionis adopted Intelligencia AI’s Portfolio Optimizer as an external PTRS estimator. This provided an independent, data-driven reference point that could be compared directly with internal PTRS calculations.
Beyond the headline PTRS values, the associated PTRS drivers proved especially impactful. These drivers offered transparency into the factors influencing success probabilities and addressed common concerns about opaque or black-box models. By examining patterns across both Ionis programs and external comparators, the team gained clearer insight into how operational factors such as trial feasibility and patient recruitment, along with scientific novelty and development complexity, materially influenced outcomes.
These insights prompted refinements to Ionis’s internal PTRS framework. Operational risks that had previously been embedded within broader qualitative scores were elevated, separated into more explicit components, and defined with clearer criteria. This ensured that feasibility and execution risks were assessed more deliberately and consistently alongside scientific considerations.
A Collaborative Thought Partnership
The engagement extended beyond platform usage. Regular discussions between Ionis and Intelligencia AI focused on PTRS methodology, benchmark design, and process integration. These conversations supported adoption, clarified interpretation of outputs, and helped align the tools with Ionis’s internal decision-making cadence.
The platform itself was described as intuitive and easy to explore, allowing users to quickly build familiarity and confidence. When questions arose, responses were timely and often led to broader refinements or tailored outputs that addressed emerging analytical needs.
The Outcome: From Static Artifacts to a Living Decision Aid
The combined use of Dynamic Benchmarks, the Portfolio Optimizer, and ongoing collaboration has produced tangible changes in how Ionis approaches PTRS and portfolio strategy.
Benchmarks are now more closely aligned with the specific modalities, targets, and development paths under discussion. Time windows reflect the regulatory and technological context most relevant to leadership decisions. Assumptions and comparators are easier to explain, test, and adjust as questions arise.
Rather than serving as static artifacts, PTRS benchmarks have become a living decision aid that evolves with the portfolio. The portfolio planning team can respond in real time to scenario-based questions, explore alternative assumptions collaboratively, and ensure that discussions are grounded in shared evidence.
Importantly, insights from PTRS drivers have reshaped internal thinking about risk. Operational feasibility and recruitment are now recognized as major, independent contributors to success, not secondary considerations. Given that changes in PTRS can translate into substantial shifts in expected future value, this more comprehensive view of risk has meaningful financial implications.
“Intelligencia AI has enabled me to feel confident discussing PTRS and excited to discuss PTRS. I see it as my trusty sidekick, and it helps me be even more successful in my role.”
— Ionis Portfolio Planning Team
The team’s role in PTRS discussions has also evolved. With greater transparency and control over the underlying analyses, they are able to engage more confidently and constructively with stakeholders. Conversations focus less on defending numbers and more on jointly understanding trade-offs, assumptions, and strategic choices.
Over time, this has contributed to stronger internal ownership of PTRS. The portfolio team designs and interprets benchmarks themselves, understands the drivers behind success probabilities, and can clearly articulate the rationale behind recommendations. PTRS has shifted from something delivered by external sources to a capability led and continuously refined within Ionis.
Looking ahead, there is interest in expanding access to these tools beyond the core portfolio planning group, enabling a broader set of stakeholders to engage directly with benchmarks and drivers. This would further support a shared, dynamic understanding of how scientific, operational, and strategic decisions influence the likelihood of success across the portfolio.
By replacing static, generic benchmarks with tailored, transparent, and responsive analytics, Ionis has strengthened its ability to make informed investment decisions in a complex and rapidly evolving clinical landscape.
For more tailored, data-rich insights, let’s talk.
