# Plain-Language Summary

**What is this about?**
Eosinophilic esophagitis (EoE) is a chronic allergic disease, affecting about 1 in 2,000 people in Western countries, in which the immune system reacts to food and inflames the esophagus — the tube that carries food to the stomach. Current treatments such as swallowed steroids and strict elimination diets can suppress the symptoms, but they do not retrain the immune system to stop over-reacting, so they must be continued for life. This work aims at that missing piece: inducing lasting tolerance to the specific foods that trigger the disease.

**What did we do?**
Using public data and computer methods — no laboratory experiments — we designed a therapy platform aimed at the three most common EoE food triggers: dairy, wheat, and soy. We used prediction tools to find the exact food-protein fragments the immune system displays, used AI structure prediction to model them in place, and then designed three delivery formats, including a tiny 20-nanometre iron-oxide nanoparticle that carries five copies of the fragment to nudge the immune system toward calming "regulatory" cells rather than attacking. We also drafted a full, costed 18–24 month preclinical plan.

**What did we find?**
We identified and modelled one strong fragment for each of the three foods (dairy, wheat, and soy), all matched to a common HLA immune type, with predicted binding strengths in a favourable range (about 17–50 nM). A notable detail: the best dairy fragment comes from the milk protein's precursor form and binds roughly 90 times more tightly than the mature version — a difference that matters for picking the right target. The AI-predicted structures scored well, with each fragment sitting correctly in its display groove, and the nanoparticle design uses standard, manufacturable chemistry expected to boost the effect about 24-fold.

**How sure are we?**
This is early, computer-based work produced during a one-week hackathon. The fragments and structures are predictions and model estimates, not experimental results, and the therapy has not been made or tested in the lab. Our designs so far cover a single common HLA type, so a major next step — flagged as an early regulatory requirement — is extending to other HLA types so more patients could benefit. The detailed preclinical roadmap sets out exactly which laboratory and safety studies would be needed next.

**Why does it matter?**
Instead of suppressing EoE indefinitely, this platform aims to retrain the immune system toward durable tolerance to specific foods. It also demonstrates that a patient-researcher living with the disease, working with AI tools, can produce a rigorous, transparent, partner-ready drug-design package quickly — a citizen-science model for tackling under-served allergic and orphan diseases.