A Systematic Evaluation of Molecular Mixture Behavior Prediction

• URL: http://arxiv.org/abs/2605.29698v1
• Date: Thu, 28 May 2026 09:58:59 GMT
• Title: A Systematic Evaluation of Molecular Mixture Behavior Prediction
• Abstract summary: We propose an evaluation framework that decomposes mixture-property error into pure-compound and interaction (non-ideal) components.<n>We show that strong absolute accuracy can mask poor recovery of non-ideal mixture behavior.<n>These results identify transfer to unseen molecules as a central challenge in molecular mixture machine learning.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Machine learning for molecular property prediction has focused largely on pure compounds, even though many practical applications depend on mixtures with intermolecular interactions. Recent work has expanded the availability of mixture datasets, but evaluation still focuses mainly on absolute accuracy. However, absolute errors in mixtures conflate pure-component contributions with deviations from ideal mixing. We propose an evaluation framework that decomposes mixture-property error into pure-compound and interaction (non-ideal) components. The framework combines leakage-aware split protocols, ideal-mixture baselines, and excess-property metrics. To support reproducible benchmarking, we curate seven matched pure and mixture physicochemical property datasets. Across multiple mixture-property tasks and model families, we find that strong absolute accuracy can mask poor recovery of non-ideal mixture behavior, and that performance drops substantially under strict molecule splits. These results identify transfer to unseen molecules as a central challenge in molecular mixture machine learning and motivate evaluation beyond absolute accuracy alone.

Related papers

• Beyond Predictive Accuracy: A Reliability-Aware Audit of Molecular Representations for Human Olfaction [0.0]
  We present a reliability-aware audit of generic molecular representations for human olfaction across four claims.<n>Human three-attribute rating geometry, based on intensity, pleasantness, and familiarity, is reproducible across participant splits.<n>MoLFormer provides no clear incremental predictive value beyond a combined RDKit-Morgan baseline in either single-molecule dataset.
  arXiv  Detail & Related papers  (2026-07-25T03:38:14Z)
• ChemFlow:A Hierarchical Neural Network for Multiscale Representation Learning in Chemical Mixtures [10.457130454437696]
  ChemFlow is a novel hierarchical framework that integrates atomic, functional group, and molecular-level features.<n>ChemFlow excels at predicting concentration-dependent properties and significantly outperforms state-of-the-art models.
  arXiv  Detail & Related papers  (2026-03-03T09:54:34Z)
• AROMMA: Unifying Olfactory Embeddings for Single Molecules and Mixtures [1.9872740012524774]
  We propose AROMMA, a framework that learns a unified embedding space for single molecules and two-molecule mixtures.<n>AROMMA achieves state-of-the-art performance in both single-molecule and molecule-pair datasets, with up to 19.1% AUROC improvement.
  arXiv  Detail & Related papers  (2026-01-27T12:54:31Z)
• Learning Continuous Solvent Effects from Transient Flow Data: A Graph Neural Network Benchmark on Catechol Rearrangement [0.0]
  This work introduces the textbfCatechol Benchmark, a high- throughput transient flow chemistry dataset.<n>We evaluate various architectures under rigorous leave-one-solvent-out and leave-one-mixture-out protocols to test generalization to unseen chemical environments.<n>We propose a hybrid GNN-based architecture that integrates Graph Attention Networks (GATs) with Differential Reaction Fingerprints (D RFP) and learned mixture-aware encodings.
  arXiv  Detail & Related papers  (2025-12-22T16:19:01Z)
• CheMixHub: Datasets and Benchmarks for Chemical Mixture Property Prediction [0.0]
  CheMixHub is a holistic benchmark for molecular mixtures.<n>It covers a corpus of 11 chemical mixtures property prediction tasks, from drug delivery formulations to battery electrolytes.<n>CheMixHub introduces various data splitting techniques to assess context-specific generalization and model robustness.
  arXiv  Detail & Related papers  (2025-06-13T21:19:27Z)
• Hierarchical Matrix Completion for the Prediction of Properties of Binary Mixtures [3.0478550046333965]
  We introduce a novel generic approach for improving data-driven models. We lump components that behave similarly into chemical classes and model them jointly. Using clustering leads to significantly improved predictions compared to an MCM without clustering.
  arXiv  Detail & Related papers  (2024-10-08T14:04:30Z)
• The Benefits of Mixup for Feature Learning [117.93273337740442]
  We first show that Mixup using different linear parameters for features and labels can still achieve similar performance to standard Mixup. We consider a feature-noise data model and show that Mixup training can effectively learn the rare features from its mixture with the common features. In contrast, standard training can only learn the common features but fails to learn the rare features, thus suffering from bad performance.
  arXiv  Detail & Related papers  (2023-03-15T08:11:47Z)
• Harnessing Hard Mixed Samples with Decoupled Regularizer [69.98746081734441]
  Mixup is an efficient data augmentation approach that improves the generalization of neural networks by smoothing the decision boundary with mixed data. In this paper, we propose an efficient mixup objective function with a decoupled regularizer named Decoupled Mixup (DM) DM can adaptively utilize hard mixed samples to mine discriminative features without losing the original smoothness of mixup.
  arXiv  Detail & Related papers  (2022-03-21T07:12:18Z)
• Quantum Coherent States of Interacting Bose-Fermi Mixtures in One Dimension [68.8204255655161]
  We study two-component atomic gas mixtures in one dimension involving both bosons and fermions. We report a rich variety of coherent ground-state phases that vary with the intrinsic and relative strength of the interactions.
  arXiv  Detail & Related papers  (2021-10-26T17:52:37Z)
• Federated Learning of Molecular Properties in a Heterogeneous Setting [79.00211946597845]
  We introduce federated heterogeneous molecular learning to address these challenges. Federated learning allows end-users to build a global model collaboratively while preserving the training data distributed over isolated clients. FedChem should enable a new type of collaboration for improving AI in chemistry that mitigates concerns about valuable chemical data.
  arXiv  Detail & Related papers  (2021-09-15T12:49:13Z)
• Unassisted Noise Reduction of Chemical Reaction Data Sets [59.127921057012564]
  We propose a machine learning-based, unassisted approach to remove chemically wrong entries from data sets. Our results show an improved prediction quality for models trained on the cleaned and balanced data sets.
  arXiv  Detail & Related papers  (2021-02-02T09:34:34Z)
• Consistent Estimation of Identifiable Nonparametric Mixture Models from Grouped Observations [84.81435917024983]
  This work proposes an algorithm that consistently estimates any identifiable mixture model from grouped observations. A practical implementation is provided for paired observations, and the approach is shown to outperform existing methods.
  arXiv  Detail & Related papers  (2020-06-12T20:44:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.