When to measure, what to measure, and in whom, determines how much can ultimately be learned from a study. Quercus Insights assists in the design of experiments so that the resulting data are as informative as possible for the scientific and development questions of interest.

This typically includes:

• assessing structural and practical identifiability of model parameters under proposed protocols

• highlighting which endpoints and timepoints are most informative

• outlining design modifications that improve downstream interpretability and model calibration

The outcome is a documented set of design recommendations and rationale that can be used to refine protocols and plan subsequent modeling efforts.

In addition to improving the interpretability of results, efficient experiment design can reduce unnecessary use of animals and align with emerging regulatory guidance that encourages more targeted, mechanism-informed preclinical work. 

Many programs involve more than one agent, and there are often numerous plausible ways to combine or sequence them. We help structure and analyze this strategy space using mechanistic models, available data, and targeted literature integration.

Typical activities may include:

• formulating explicit hypotheses about mechanisms of interaction or resistance

• exploring candidate dosing and sequencing strategies within a mechanistic or semi-mechanistic framework

• prioritizing a small number of strategies for further experimental evaluation

The objective is to provide a transparent, model-informed basis for selecting and testing combination or sequencing options.

Patient heterogeneity is central to efficacy, safety, and trial design. Quercus Insights develops and analyzes virtual populations (ensembles of model parameter sets representing plausible biological variability, both theorized, or seeded from digital twins if such data are available) to explore how different subgroups may respond to a given intervention.

Depending on project needs, this work can include:

• constructing virtual cohorts consistent with available data and biological knowledge

• characterizing response patterns across parameter and covariate space

• generating candidate stratification or enrichment strategies to be evaluated against internal or external datasets
  
  

The intent is not to replace clinical insight, but to provide a structured, mechanistic perspective on heterogeneity that can inform both analysis and design.

Large or highly detailed models can be difficult to maintain, communicate, and apply in a development setting. We work with existing models, whether developed internally, by external partners or found in the literature, to simplify them while retaining the behaviors that are critical for specific questions.

This may involve:

• identifying and removing non-essential states or processes

• applying timescale or structural reductions

• comparing the reduced and original models across relevant scenarios to assess robustness
  
  

The deliverable is a simplified, documented model supported by comparative analyses, making it easier for teams to reuse and scrutinize the model in future work.

In some cases, existing quantitative methods are not well aligned with the questions or data structure at hand. Quercus Insights can design and implement tailored analytical or modeling approaches for specific projects.

From the outset, we work with clients to distinguish between:

• generalizable methodology that may be suitable for subsequent publication

• project-specific implementation that is developed under the client’s confidentiality and ownership requirements

We are committed to advancing methodological practice where possible, while respecting data privacy and contractual constraints. Publication or internalization of methods can be addressed explicitly in the engagement terms.

Not every situation requires a full modeling project. Often, there is a focused question, unexpected data behavior, or a need to decide whether quantitative modeling would add value. For such cases, we offer limited-scope consultations aimed at:

• clarifying the problem and decision context

• outlining plausible mechanistic explanations and testable hypotheses

• suggesting candidate experiments, analyses, or modeling approaches

These sessions are designed as a low-commitment way for teams, especially start-ups without internal modeling groups, to explore whether deeper collaboration would be useful.

Early-stage organizations may have strong biology and clinical capabilities, but limited senior modeling leadership. Quercus Insights can provide fractional support in that capacity for a defined period or project.

Depending on needs, this can include:

• helping to frame quantitative questions and prioritize effort

• advising on model structure and choice of methods

• reviewing and stress-testing internal or external modeling work

• supporting communication of quantitative results to broader teams and stakeholders

The emphasis is on building internal capability and clarity.

In addition to project work, Quercus Insights can provide invited talks or small workshops on topics including but limited to experiment design and identifiability, mechanistic thinking for complex disease, combination and sequencing strategies, virtual populations, and bell-shaped dose–response. These can be tailored to scientific, quantitative, or mixed audiences.