Drug Discovery · Disease Dynamics

Causal
Dynamical Systems
Drug Discovery

We model disease as a dynamic system — and explore when interventions can shift its state.

DISCUSS THE APPROACH
Scroll to explore
01
The foundation

Disease is a
dynamical system

Feedback loops stabilize pathology
Thresholds determine reversibility
Intervention outcomes depend on system state
Most approaches model data. We model the biological system that generates disease progression .
02
The model

A causal model of
disease progression

Molentio encodes disease as a system of interacting biological processes, connected through their causal structure:

Senescence→
SASP→
Inflammation→
Myofibroblasts→
Fibrosis→
Organ function
Feedback structure Pathological states reinforced by internal loops
Nonlinear transitions State changes that cannot be predicted from averages
Stable disease states Multiple attractors with distinct clinical outcomes

Disease progression is not linear — it is state-dependent.

03
The core insight

The reversibility
boundary

Disease systems exhibit multiple stable states. Between them lies a boundary — the point beyond which a given intervention is insufficient to return the system to a healthy state.

We aim to determine whether a system state can be shifted across this boundary, and the conditions required to do so.

HEALTHY FIBROTIC REVERSIBILITY BOUNDARY patient

Two stable disease states separated by a potential barrier. The teal basin is the healthy attractor; the amber basin is the fibrotic attractor. The patient marker sits within the fibrotic basin, but close enough to the boundary that a well-targeted intervention may still shift the system back to the healthy state.

Therapy succeeds only if the system can still be shifted.

04
Applications

From state
to strategy

Molentio translates mechanistic disease models into actionable insights for drug development.

01
Patient Stratification

Identify which disease states are theoretically reversible under a given intervention.

02
Intervention Design

Determine whether a target has sufficient leverage to shift system dynamics.

03
Combination Strategy

Identify compensatory pathways and required multi-target interventions.

04
Trial Design

Align intervention mechanism with disease state to improve signal detection.

05
State inference

Inferring the
underlying state

Clinical measurements provide a partial view of disease. Molentio focuses on estimating the underlying mechanistic state that drives progression.

By combining observed data with model constraints, we determine where a patient's system lies — and which interventions remain viable.

What matters is not what we observe — but the state the system is in.

06
Initial indication
Chronic Kidney Disease

Senescence-driven
fibrosis

Our initial work focuses on chronic kidney disease, where fibrosis progression exhibits clear dynamical structure — making it an ideal validation ground for causal dynamical modeling.

Stable pathological states identified and explored
Intervention regimes mapped to system dynamics
Reversibility conditions characterized

This serves as the foundation for expansion into broader fibrotic diseases.

07
Differentiation

Why Molentio

Causal not correlational
Interventions are modeled as actions on the system, not inferred from statistical associations.
Dynamical not static
Disease is treated as a time-evolving process with internal structure — not a snapshot.
Mechanistic not black-box
Every prediction is grounded in biology. The model is interrogable at every step.
State-dependent not average-based
Outcomes depend on where the patient's system is — not the population mean.
Collaboration

Collaborate
with Molentio

We engage with biotech and pharma teams to explore how causal dynamical models can be applied to drug discovery and development.

Mechanistic Disease Modeling Mechanistic Regime Analysis Causal Target Discovery Disease Model Expansion

Each collaboration aims to generate program-specific, mechanistically grounded insights for therapeutic development.

Explore collaboration