Ruth-Anne Pai, PhD
Manuscripts

Three manuscripts, from design to evidence

A personalized antigen-selection engine for pMHC-II tolerance in EoE; a target-and-binder brief on the CCL26 / POSTN effector axis; and a dual-arm de novo design campaign for esophageal cancer that shows the same engine generalizes to a second disease. All three are computational preprints (not peer reviewed), openly documented, with structures and data available to download.

Manuscript 1

A patient-specific antigen-selection engine for personalized pMHC-II tolerance therapy in eosinophilic esophagitis

Ruth-Anne Pai, PhD · immunologist and person living with EoE · produced with Claude Science (AI) support

Rather than broadly suppressing the immune system, this work re-educates the specific food-reactive T-cell clones that drive EoE — and it computes which epitopes matter for a given person. Because presented-epitope load is strongly HLA-dependent, antigen priority is per-patient. The engine scans the dominant dairy, wheat, and soy allergens across common HLA-DRB1 alleles, then hands a folded, groove-validated lead to a tolerogenic nanoparticle (PACT: one shared backbone, a swappable peptide cassette).

4,640
peptide × allele evaluations
832
strong binders (IC₅₀ < 500 nM)
13–352
strong binders per allele (27× range)
≥0.87
ipTM structural gate, full in-groove
Decision tree showing how the PACT engine routes an individual EoE patient: from diagnostic biopsy and blood through HLA-II typing and trigger profiling to one of three manufacturing tiers, all converging on an N-of-1 nanoparticle product Decision tree showing how the PACT engine routes an individual EoE patient: from diagnostic biopsy and blood through HLA-II typing and trigger profiling to one of three manufacturing tiers, all converging on an N-of-1 nanoparticle product
The engine, per patient. Every patient is manufactured N-of-1: HLA-II typing plus trigger profiling routes their {{HLA × food-epitope}} pair to a validated component (Tier A), a computational prior requiring an ex-vivo functional gate (Tier B), or a full characterization workup (Tier C). One shared backbone, a swappable peptide cassette (PACT). The evidence tier sets the route — not whether a patient is served.
Two panels: left, a curve showing qualified-backbone HLA coverage rising from 24% with one backbone to over 97% with fifteen; right, a bar chart of EoE food triggers with dairy the most common and the tetramer-validated anchor targeting it directly
Coverage that compounds. Because presented-epitope load is strongly HLA-dependent (13–352 strong binders per allele, a 27-fold range), antigen priority is per-patient. A small qualified-backbone library reaches most patients (~78% at 5 backbones, ~90% at 8), while N-of-1 manufacturing reaches everyone from day one. Dairy leads the trigger distribution and is the tetramer-validated anchor.
Epitope discovery: affinity ranking, IEDB percentile stringency, protein-context dependency, and validation status
Epitope discovery & validation. Affinity ranking, IEDB stringency, and the ~92-fold context dependency that favored the dairy precursor form. The dairy epitope is tetramer-validated; wheat and soy are computational priors flagged for Phase-1 functional validation.

Structural models

Three pMHC-II:peptide complexes were co-folded (ESMFold2-Fast, GPU) and inspected for canonical groove geometry. Download the atomic coordinates below — .pdb files open in PyMOL, ChimeraX, or any Mol* viewer.

ComplexPeptide (15-mer)ipTMStructure
Dairy pMHC-IIKIHPFAQTQSLVYPF0.872dairy_pmhc.pdb
Wheat pMHC-IIIHNVVHAIILHQQQQ0.896wheat_pmhc.pdb
Soy pMHC-IIAYPFVVNATSNLNFL0.891soy_pmhc.pdb

Preclinical roadmap

Preclinical development roadmap: five phases over an 18-24 month timeline to IND submission
A five-phase, 18–24 month path to IND submission — human ex-vivo validation, murine proof-of-concept, GLP toxicology and biodistribution, GMP manufacturing, and regulatory filing — each with explicit go / no-go decision gates.
Manuscript 2

De novo design of neutralizing protein binders against CCL26 and POSTN, two convergent targets in EoE

Ruth-Anne Pai, PhD · immunologist and person living with EoE · produced with Claude Science (AI) support

CCL26 and POSTN emerged from a nine-cohort transcriptomic meta-analysis as high-evidence, no-trial biology — the eosinophil-recruitment axis and the barrier-remodeling axis. From 80 RFdiffusion backbones → 1,920 SolubleMPNN sequences → 60 Boltz-2-assessed complexes, lead binders engage the intended functional epitopes: CCL26 ipTM 0.942 (4/5 hotspots, 1,220 Å² buried) and POSTN ipTM 0.901 (pLDDT 0.911, 3/4 hotspots).

Structural render of a designed binder (orange) engaging the CCL26 chemokine surface (grey), with the target epitope highlighted in blue
CCL26 (eotaxin-3) binder. A designed binder (orange) occludes the CCR3-engagement surface of the mature chemokine (epitope in blue) — neutralizing the dominant EoE eosinophil chemoattractant.
Structural render of a designed helical binder (orange) engaging the POSTN target surface (grey), interface residues in blue
POSTN (periostin) binder. A helical binder engaging the POSTN interface — targeting the matricellular protein that amplifies eosinophil adhesion and tissue remodeling in EoE.

Design rationale

Two routes for CCL26

A neutralizing anti-CCL26 IgG1 must occlude the N-loop / 40s-loop docking surface to block CCR3 site-1 engagement — a high-affinity (KD < 1 nM) requirement for the small, 71-aa antigen. Alternatively, a small-molecule CCR3 antagonist targets the invariant receptor and captures the other eotaxins (CCL11/CCL24) on the same axis for a broader anti-eosinophil effect.

ESM-guided escape hardening

The sole ESM escape-risk position in the CCL26 epitope (H39) sits inside the N-loop docking surface — so the design co-engages the disulfide-rigid, conserved 40s loop rather than relying on H39. For the IL1RL1 interface, CDRs are steered toward conserved aromatic hotspots (Y119 / F245) and away from the escape-prone rim.

Why it differentiates

Effector-specific eosinophil-recruitment blockade — narrower and more targeted than dupilumab's broad Th2 blockade, and universal across EoE endotypes.

The honest result: a negative control that matters

We ran 36 scrambled and random decoy complexes through the same pipeline. 92–100% of decoys cleared the same ipTM/pLDDT gate that 59 of 60 designs cleared — so a high structural pass rate is close to the null and, on its own, does not demonstrate a real binder. What separates the leads is focal, on-epitope engagement, not the confidence score. Claude Science flagged this over-claim during the work; reporting it is part of the point.

Manuscript 3

Steering into competitive whitespace: a public-data campaign nominates GUCY2C and DKK1 and delivers de novo binder leads across the esophageal-adenocarcinoma trajectory

Ruth-Anne Pai, PhD · immunologist and person living with EoE · produced with Claude Science (AI) support

The strongest test of a platform is whether it works on a disease it wasn't built for. I pointed the same engine at esophageal adenocarcinoma (EAC) — a copy-number-driven cancer at the end of the Barrett's trajectory — and asked it to find tractable targets in competitive whitespace, then design binders. It nominated a dual-arm program: GUCY2C for a T-cell-engager treatment arm, and DKK1 for a neutralizing-trap interception arm in high-risk Barrett's.

182
TCGA EAC tumors profiled (cBioPortal)
351
EAC trials mapped (106 active)
16
de novo binders (8 per target)
15/16
clear the ipTM > 0.5 interface line
Bar chart of EAC genomic alterations from TCGA (n=182): TP53 mutated in 87%, CDKN2A deleted in 39%, CCND1 amplified in 35%, with additional amplified surface receptors and angiogenic factors
EAC is copy-number-driven. TCGA PanCancer (n=182): TP53 altered in 87%, CDKN2A deep deletion 39%, CCND1 amplification 35%. The drug-tractable drivers are amplified surface receptors and secreted factors — which is what steered target selection toward a surface antigen and a secreted ligand.
Scatter map of EAC targets plotting clinical trial crowding against development maturity; GUCY2C and Barrett's interception sit in open whitespace while HER2 and PD-1 are crowded and VEGF is mature
Competitive whitespace. Novel surface antigens (GUCY2C, B7-H3, TROP2) and precursor interception are open; HER2 and PD-1/PD-L1 are crowded, VEGF is mature. GUCY2C and Barrett's interception occupy the least-contested biology.
Two-panel figure: (A) interface ipTM versus complex pLDDT for 16 designed binders against GUCY2C and DKK1, with gucy2c_bb2 and dkk1_bb1 as top leads; (B) ranked bar chart showing 15 of 16 designs clear the ipTM 0.5 interface line
De novo binder design (Stage 6, GPU). RFdiffusion → ProteinMPNN → Boltz-2 fold-back. Top leads: GUCY2C gucy2c_bb2 (ipTM 0.915; flagged for a high-alanine developability liability) and DKK1 dkk1_bb1 (ipTM 0.858, pLDDT 0.908 — the cleanest design in the set). Reported as pilot-scale in-silico leads: ipTM is a docking-confidence proxy, not a measured affinity.

Treatment arm — GUCY2C

A gut-restricted surface antigen (GTEx: ~85% GI-restricted expression) nominated #1 in 80.7% of 20,000 weight-perturbation draws, with CDH17 as a backup. Designed for a T-cell-engager format. Open question: the normal-gut therapeutic window.

Interception arm — DKK1

A secreted Wnt modulator targeted with a neutralizing trap for high-risk Barrett's — interception before invasive cancer. Open question: a dysplasia-regression surrogate endpoint that a trial could actually read out.

Program verdict: TRACTABLE (0.87). The two make-or-break questions above are named as open, not claimed. This is a design-stage dossier, not a clinical result.