Before damage begins.
Which changes in diabetes could set nerve injury in motion? We trace the steps from an initial exposure to physical damage and ask where prevention could make a difference.
Our research
We use AI to build an interpretable, quantitative understanding of biology, from the mechanisms that sustain a neuron to the changes that cause disease.
Understand each link,
test the whole explanation.
The DPN mechanism explorer presents mechanisms under investigation in diabetic peripheral neuropathy. We begin with a precise question about nerve damage, break the proposed explanation into its necessary biological steps, and ask what evidence would support or overturn each one.
An explanation needs to hold together across the same cells, conditions, and timescale. We trace how changes in molecules and cellular processes could lead to physical nerve loss, accounting for alternative routes and the body’s capacity to compensate.
Our current work combines public biological evidence with computational experiments. The goal is to develop mechanistic explanations that can be inspected, challenged, and used to make meaningful predictions about prevention and repair.
Publications by our affiliated researchersOne disease, different opportunities to help.
Prevention, slowing established injury, and repair each ask a different question. We define the outcome clearly and follow the evidence relevant to it.
Protect what remains.
What keeps injury progressing, and what helps a nerve withstand it? We examine the balance between damage, cellular maintenance, and the capacity to compensate.
Make recovery possible.
What would it take for damaged nerves to recover? We investigate how new nerve endings could grow, remain connected, and restore lasting function.
Clear reasoning,
measurable predictions.
AI helps us examine literature, analyze data, and build models. We use that capacity to make causal reasoning explicit and run calculations that other researchers can inspect and challenge.
- 01
Define the claim
State exactly what might cause what, in which cells or people, and over what timescale. Identify the measurable change that would make the answer meaningful.
- 02
Decompose the mechanism
Break the explanation into its necessary biological steps. Map alternative routes, interactions, and what must hold true for the whole argument to work.
- 03
Quantify and test
Use data and computation to estimate rates, effects, and uncertainty. Check physical constraints and test predictions that distinguish competing explanations.
- 04
Revise with the evidence
Update the full explanation as each result arrives. Make remaining uncertainty explicit and pursue the next test that could meaningfully change our understanding.
Following the life of a neuron.
Our questions connect molecular events to the long-term survival and function of a nerve. We ask how much a process changes, where it acts, and whether that change is sufficient to matter.
Inputs and exposure
What reaches the neuron from its surroundings, at what concentration, and for how long? We examine whether the proposed exposure could occur under the conditions of diabetes.
Energy and maintenance
Can a nerve meet its energy needs, move essential materials, and replace damaged components? We study the rates and limits that determine whether maintenance can keep up.
Damage and compensation
Which changes cause physical injury, and which reflect a response to it? We test how protective processes, supporting cells, and alternative routes affect the outcome.
Preservation and repair
Would changing a proposed mechanism preserve nerve endings or support durable replacement? We connect local effects to their magnitude, persistence, and relevance to people.
Confidence that follows
the evidence.
We seek the strongest answer the evidence can support, with a clear account of what remains uncertain and what could change our minds.
We distinguish measured effects, model predictions, and judgments about whether an explanation is true. When we assign probabilities, we make the assumptions visible and show how different reasonable interpretations of the evidence change the assessment.
Connected biological steps and shared evidence are considered together. Support for one part of a mechanism is evaluated in the context of the full causal argument, including its unresolved links.
Our aim is to make confidence track reality as closely as possible. A result may support a claim, contradict it, or leave it unresolved. We explain which conclusion is warranted and keep looking for the measurement or calculation that could resolve the next uncertainty.
Our work is currently computational and exploratory. Translating a promising mechanism into a treatment requires experimental and clinical evidence.
Better futures start
with better questions.
Join a coalition working to understand biology and build a future without neurological disease.
Get in touch