Rather than beginning with a preferred modeling framework, we start with the specific decision context and practical limitations.

In practice, this means clarifying:

• what decisions are actually on the table (e.g., which experiment to run next, how to sequence agents, how to prioritize mechanisms)

• what time, budget, and data are realistically available

• what level of uncertainty is acceptable at this stage of the program

The goal is to align quantitative work with what the team must decide in the near term, rather than to build models for their own sake.

Many downstream difficulties in modeling and interpretation trace back to how studies were designed. A substantial part of the work therefore focuses on making proposed or planned experiments as informative as possible.

Typical elements include:

• assessing whether planned measurements and schedules can reasonably constrain key mechanisms or parameters

• suggesting alternative or additional timepoints, endpoints, or cohorts where they would significantly improve interpretability

• highlighting trade-offs between complexity of design and the ability to draw quantitative conclusions later

This emphasis on design is particularly important for early-stage programs with limited opportunities to generate data.

Thoughtful design is not only analytically stronger; it can also make more respectful use of experimental resources, including animals, by focusing on measurements that genuinely inform key questions. 

When we develop or adapt mechanistic models, we do so with specific questions and data structure in mind, rather than adhering to a fixed “house model”.

This often entails:

• choosing a level of biological and mathematical detail that is appropriate for the available information

• making explicit which mechanisms are included, which are simplified, and which are held in reserve as hypotheses

• using virtual populations or scenario analyses where they add insight into robustness, variability, or translation, without overstating predictive power.

The intent is to have models that are transparent and discussable, not black boxes.

A recurring theme in the work is making assumptions and methods visible so they can be scrutinized and reused.

Accordingly, projects are structured to provide:

• documented model structures, with clear notation and rationale for key equations and mechanisms

• code that has been reviewed and annotated for clarity, so that internal teams can work with it independently

• written summaries that connect quantitative outputs back to the original questions and constraints

This is particularly important for start-ups and small teams who may need to rely on the same models and methods over a longer horizon.

Quercus Insights is not a contract research organization and does not aim to replicate that model. The practice is intentionally small and designed to complement internal teams rather than to absorb large volumes of work.

In contrast to many CRO-style engagements, we place emphasis on:

• scoping work tightly around a defined set of decisions and questions

• building in a limited amount of follow-up to revisit conclusions when new data become available

• treating iteration and clarification as part of the scientific process, not as automatic change orders

• avoiding unnecessary expansion of model scope or complexity when a more focused structure is sufficient

These principles are reflected in how projects are proposed, executed, and documented.

Most engagements are collaborative and iterative rather than transactional. Teams are encouraged to bring not only data, but also their hypotheses, concerns, and constraints.

A typical collaboration may involve:

• an initial phase of problem framing and review of existing work and data

• one or more focused blocks of quantitative work (design support, modeling, method development)

• structured mutually agreed upon check-ins to discuss results, limitations, and possible next steps

Where projects would benefit from additional expertise, i.e., specialized data analysis or alternative stratification frameworks, we can work with a small network of collaborators, keeping the core engagement focused and coordinated.