{
  "abstract": "Molecular graph encoding often relies on a single, static structure, ignoring the thermodynamic ensemble of molecules that are present in solution. Here, we introduce EnsembleEGNN, a foundation model that encodes structural ensembles by processing individual conformers through shared equivariant graph neural network layers, pooled with a set attention block, to make property predictions from the whole ensemble. Pretrained on the CREMP cyclic peptide dataset using multi-task self-supervision, the model is trained to encode the conformational variability of each molecule. When predicting membrane permeability from the CycPeptMPDB benchmark, EnsembleEGNN achieves an $R^2$ of $0.477$ under random cross-validation, outperforming a sequence-only BERT baseline ($R^2=0.439$). This representation advantage persists under rigorous out-of-distribution Butina splits ($R^2=0.401$ versus $0.354$). Finally, a hybrid architecture co-training EnsembleEGNN with the BERT model achieves the highest overall accuracy across both random ($R^2=0.538$) and structural holdout evaluations ($R^2=0.444$). These results demonstrate that encoding conformational ensembles into latent representations improves predictions for properties governed by thermodynamics.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2607.21561",
  "author_total": 3,
  "authors": [
    {
      "name": "Aaron L. Feller"
    },
    {
      "name": "Kris Deibler"
    },
    {
      "name": "Maxim Secor"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/dx32cw6bcnqap.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2607.21561",
  "fulltext_url": null,
  "id": "dx32cw6bcnqap",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2607.21561",
  "language": null,
  "licence": "CC-BY-4.0",
  "licence_label": "CC BY 4.0",
  "licence_source": "article",
  "licence_url": "http://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.LG"
    ],
    [
      "arxiv",
      "q-bio.BM"
    ],
    [
      "openalex",
      "1703|Computational Drug Discovery Methods"
    ]
  ],
  "openalex": "W7170883142",
  "pdf_url": "https://arxiv.org/pdf/2607.21561",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-30",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/dx32cw6bcnqap.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:25:52Z",
      "native_id": "2607.21561",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2607.21561"
    }
  ],
  "subjects": [
    {
      "field": "biochemistry",
      "origin": "arxiv:q-bio.BM",
      "subfield": "biophysics"
    }
  ],
  "title": "EnsembleEGNN: Set-Based Graph Learning for Thermodynamic Ensembles of Cyclic Peptides",
  "title_text": "EnsembleEGNN: Set-Based Graph Learning for Thermodynamic Ensembles of Cyclic Peptides",
  "type": "preprint",
  "updated_at": "2026-10-30T12:25:52Z",
  "url": "https://research.draconitic.com/paper/dx32cw6bcnqap",
  "venue": "arXiv",
  "versions": []
}