Molecular Representation Learning

Overview

Molecular representation learning studies how to encode molecules and biomolecules so models can recognize chemical structure, predict properties, and support tasks such as drug screening and molecular design. Approaches range from fingerprints and molecular graphs to SMILES language models, 3D equivariant networks, fragment-based methods, and multimodal models combining structure with spectra, text, or biological measurements. Self-supervised and contrastive pretraining are common ways to learn from unlabeled structures, while supervised models use task-specific labels for properties such as solubility, binding affinity, or toxicity. A central challenge is reliable transfer to new chemical scaffolds and biological settings: larger pretrained models do not always beat simpler fingerprints or descriptors, and results depend strongly on data splits, representation choices, and whether physical and chemical information is preserved.

Latest papers

219 papers12 papers in the last four weeks, down 57% on the four weeks before. 0.1% of all new papers.

Feb 19, 2026

JEPA-DNA: Grounding Genomic Foundation Models through Joint-Embedding Predictive Architectures

Genomic Foundation Models (GFMs) typically rely on Masked Language Modeling (MLM) or Next-Token Prediction (NTP) to learn the "Laws of Nature". While effective at capturing local syntax, these generative paradigms prioritize token-level reconstruction over high-level functional context. We introduce JEPA-DNA, a model-agnostic continual training framework that integrates a Joint-Embedding Predictive Architecture (JEPA) with traditional generative objectives. By supervising global sequence embeddings in a latent space, JEPA-DNA forces models to predict the functional representations of masked genomic segments, shifting the learning signal from token recovery to semantic alignment. We evaluate JEPA-DNA on 17 diverse genomic benchmark tasks, demonstrating consistent gains in linear probing and zero-shot performance regardless of the underlying GFM architecture or generative objective, and establishing a new state-of-the-art for GFMs under the linear-probing protocol. Through extensive ablation studies, we further characterize the synergistic interplay between generative and latent objectives. Our code is provided in the supplementary material.
Feb 18, 2026

Cryo-SWAN: the Multi-Scale Wavelet-decomposition-inspired Autoencoder Network for molecular density representation of molecular volumes

Learning robust representations of 3D shapes from voxelized data is essential for advancing AI methods in biomedical imaging. However, most contemporary 3D computer vision approaches operate on point clouds, meshes, or octrees, while volumetric density maps, the native format of structural biology and cryo-EM, remain comparatively underexplored. We present Cryo-SWAN, a voxel-based variational autoencoder inspired by multi-scale wavelet decomposition. The model performs conditional coarse-to-fine latent encoding and recursive residual quantization across perception scales, enabling accurate capture of both global geometry and high-frequency structural detail in molecular density volumes. Evaluated on ModelNet40, BuildingNet, and a newly curated dataset of cryo-EM volumes, ProteinNet3D, Cryo-SWAN consistently improves reconstruction quality over state-of-the-art 3D autoencoders. We demonstrate that the molecular densities organize in learned latent space according to shared geometric features, while integration with diffusion models enables denoising and conditional shape generation. Together, Cryo-SWAN is a practical framework for data-driven structural biology and volumetric imaging.
Feb 16, 2026

BindCLIP: A Unified Contrastive-Generative Representation Learning Framework for Virtual Screening

Virtual screening aims to efficiently identify active ligands from massive chemical libraries for a given target pocket. Recent CLIP-style models such as DrugCLIP enable scalable virtual screening by embedding pockets and ligands into a shared space. However, our analyses indicate that such representations can be insensitive to fine-grained binding interactions and may rely on shortcut correlations in training data, limiting their ability to rank ligands by true binding compatibility. To address these issues, we propose BindCLIP, a unified contrastive-generative representation learning framework for virtual screening. BindCLIP jointly trains pocket and ligand encoders using CLIP-style contrastive learning together with a pocket-conditioned diffusion objective for binding pose generation, so that pose-level supervision directly shapes the retrieval embedding space toward interaction-relevant features. To further mitigate shortcut reliance, we introduce hard-negative augmentation and a ligand-ligand anchoring regularizer that prevents representation collapse. Experiments on two public benchmarks demonstrate consistent improvements over strong baselines. BindCLIP achieves substantial gains on challenging out-of-distribution virtual screening and improves ligand-analogue ranking on the FEP+ benchmark. Together, these results indicate that integrating generative, pose-level supervision with contrastive learning yields more interaction-aware embeddings and improves generalization in realistic screening settings, bringing virtual screening closer to real-world applicability.
Feb 7, 2026

Learning Molecular Chirality via Chiral Determinant Kernels

Chirality is a fundamental molecular property that governs stereospecific behavior in chemistry and biology. Capturing chirality in machine learning models remains challenging due to the geometric complexity of stereochemical relationships and the limitations of traditional molecular representations that often lack explicit stereochemical encoding. Existing approaches to chiral molecular representation primarily focus on central chirality, relying on handcrafted stereochemical tags or limited 3D encodings, and thus fail to generalize to more complex forms such as axial chirality. In this work, we introduce ChiDeK (Chiral Determinant Kernels), a framework that systematically integrates stereogenic information into molecular representation learning. We propose the chiral determinant kernel to encode the SE(3)-invariant chirality product and employ cross-attention to integrate stereochemical information from local chiral centers into the global molecular representation. This design enables explicit modeling of chiral-related features within a unified architecture, capable of jointly encoding central and axial chirality. To support the evaluation of axial chirality, we construct a new benchmark for electronic circular dichroism (ECD) and optical rotation (OR) prediction. Across four tasks, including R/S configuration classification, enantiomer ranking, ECD spectrum prediction, and OR prediction, ChiDeK achieves substantial improvements over state-of-the-art baselines, most notably yielding over 7% higher accuracy on axially chiral tasks on average.
Sep 30, 2025

UncertainGen: Scalable Uncertainty-Aware Representation Learning of DNA Sequences

Learning effective representations of DNA sequences is fundamental to genomic data analysis, especially in the presence of high sequence similarity and inter-species DNA sharing. Existing approaches rely on deterministic embeddings, such as k-mer statistics or representations from large language models, which fail to capture the inherent uncertainty in short and ambiguous DNA fragments. We propose UncertainGen, a lightweight probabilistic representation learning framework that embeds DNA sequences as distributions in a latent space. The learned embedding variance captures positional ambiguity, increasing for sequences compatible with multiple genomic clusters. We establish theoretical guarantees on embedding distinguishability, and show that the probabilistic formulation induces a data-adaptive metric that expands the effective representational space. We evaluate UncertainGen in the metagenomic binning task to demonstrate the practical benefits of uncertainty-aware representations. Experiments show that probabilistic embeddings consistently outperform deterministic k-mer and large genome foundation models, while remaining lightweight and efficient.
Aug 19, 2025

A Large Scale Investigation of Scaling Limits in Chemical Language Models

Chemical Language Models (CLMs) are increasingly used in de novo drug design, driven by recent growth in model scale, compute, and dataset size. However, the relationship between design choices, training dynamics, and downstream generation quality remains poorly understood. We present a compute-controlled scaling study of CLMs comprising more than 30,000 experiments across molecular representations (SMILES, SELFIES, SAFE), tokenizations (atom-level and byte-pair encoding), model scales (0.5M-1B parameters), leakage-controlled datasets (MOSES, ChEMBL, PubChem, ZINC-22), and architectures (decoder-only and encoder-decoder). By fitting IsoFLOP profiles, we establish clear scaling trends in pretraining loss, but find that these improvements do not translate into comparable gains in goal-directed molecular design. Layer-wise probing and sparse autoencoder analysis reveal continued development of chemical representations: chemical syntax saturates early, while semantic properties emerge more slowly and become increasingly accessible with further training and model scale. These representational gains coexist with diminishing improvements in goal-directed generation under the evaluated protocols and oracle budgets. Our resulting suite of models, NovoMolGen, achieves state-of-the-art results, outperforming prior CLMs and specialized generative models in goal-directed molecular generation across drug discovery tasks. These findings expose a disconnect between chemical representation learning and downstream molecular design, motivating the development of pretraining paradigms that more directly learn chemical semantics.
Aug 8, 2025

Benchmarking Pretrained Molecular Embedding Models For Molecular Representation Learning

Pretrained molecular embedding models spanning graph neural networks, text transformers, and hybrid architectures have been widely used in drug discovery, yet no large-scale, rigorous comparison of their representations existed. We evaluated 28 models across 25 classification datasets under a controlled framework using hierarchical Bayesian Bradley-Terry statistical testing. All neural models except CheMeleon failed to outperform the classical ECFP count fingerprint; CheMeleon was the only model rated as better. However, further testing on external validation datasets, based on scaffold cross-validation and a Bayesian hierarchical correlated t-test, revealed no definite advantage of either embedding method. As a practical takeaway, we recommend evaluating both ECFP fingerprints and CheMeleon embeddings in novel QSAR/QSPR studies. Overall, in the context of all neural networks evaluated, these results also challenge the perceived progress in molecular representation learning and highlight the need for rigorous evaluation standards in the field.
Jul 3, 2025

A Machine Learning Benchmarking Framework for Lipid Nanoparticle Transfection Efficiency Prediction

The discovery of new ionizable lipids for efficient lipid nanoparticle (LNP)-mediated RNA delivery remains a major bottleneck in RNA therapeutics development. Recent advances demonstrate the potential of machine learning (ML) models to predict transfection efficiency directly from lipid structure, enabling high-throughput virtual screening and accelerating lead identification. However, as new models for LNP transfection prediction continue to emerge, the lack of rigorous and standardized benchmarking poses a significant risk and may undermine confidence in their reliability for discovery. Here, we present a robust ML benchmarking framework for evaluating transfection prediction models based on ionizable lipid structures. This framework systematically benchmarks diverse molecular representations paired with a broad range of ML architectures spanning traditional models, feedforward neural networks, and state-of-the-art graph-based methods. In addition, the presented framework supports assessment of model generalization and evaluates prediction reliability beyond standard regression metrics. Using a curated dataset of 1,100 unique ionizable lipid structures derived from the HeLa transfection dataset originally reported by Xu et al., we show that within this framework, models leveraging explicit molecular substructure encoding consistently achieve the highest predictive accuracy and should serve as essential baselines for the development of new, more sophisticated models. In contrast, some current graph-based models, including AGILE, Chemprop, and KPGT, tend to show comparatively lower accuracy. The presented framework provides a standardized, transparent, and comprehensive benchmarking resource that enables meaningful comparison of emerging architectures and establishes strong baselines for future development of predictive models in lipid-based RNA delivery.
May 29, 2025

Equivariant Spherical Transformer for Efficient Molecular Modeling

Equivariant Graph Neural Networks (GNNs) have significantly advanced the modeling of 3D molecular structure by leveraging group representations. However, message passing built only from exactly equivariant operations can have restricted practical expressiveness at a fixed representation degree. We introduce the Equivariant Spherical Transformer (EST), a plug-and-play module that applies Transformer attention to spherical spatial representations obtained from group representations. EST explicitly mixes information across sampled orientations, unlike prior orientation-wise spherical operations, while retaining a strong approximate equivariant inductive bias through near-uniform spherical sampling. As demonstrated by our experiments on challenging benchmarks like OC20, MPtrj and QM9, EST-based models achieve state-of-the-art performance. For the complex molecular systems within OC20, small models empowered by EST can outperform some larger models and those using additional data. We characterize the ideal closed-domain condition under which the continuous operation is exactly equivariant and empirically quantify the finite-sampling accuracy--equivariance--memory trade-off used in practice.
May 3, 2025

BOOM: Benchmarking Out-Of-distribution Molecular Property Predictions of Machine Learning Models

Data-driven molecular discovery leverages artificial intelligence/machine learning (AI/ML) and generative modeling to filter and design novel molecules. Discovering novel molecules requires accurate out-of-distribution (OOD) predictions, but ML models struggle to generalize OOD. Currently, no systematic benchmarks exist for molecular OOD prediction tasks. We present BOOM\mathbf{BOOM}, b\mathbf{b}enchmarks for o\mathbf{o}ut-o\mathbf{o}f-distribution m\mathbf{m}olecular property predictions: a chemically-informed benchmark for OOD performance on common molecular property prediction tasks. We evaluate over 150 model-task combinations to benchmark deep learning models on OOD performance. Overall, we find that no existing model achieves strong generalization across all tasks: even the top-performing model exhibited an average OOD error 3x higher than in-distribution. Current chemical foundation models do not show strong OOD extrapolation, while models with high inductive bias can perform well on OOD tasks with simple, specific properties. We perform extensive ablation experiments, highlighting how data generation, pre-training, hyperparameter optimization, model architecture, and molecular representation impact OOD performance. Developing models with strong OOD generalization is a new frontier challenge in chemical ML. This open-source benchmark is available at https://github.com/FLASK-LLNL/BOOM
Mar 11, 2025

PUMA: Learning a Mutation-Aware Vocabulary of Protein Units

Modeling protein sequences as a language has made language models a powerful tool in computational biology, yet the language itself remains poorly understood. A key step toward understanding it is identifying its constituent units. In natural languages, morphemes can occur in multiple forms; similarly, in proteins, mutations can give rise to variations of a unit that persist through evolution, forming families of related units. We introduce PUMA (Protein Units via Mutation-Aware Merging), an algorithm that learns protein units from sequence and explores their mutational variants using substitution matrices, forming a genealogy of unit families. Our results show that mutations remaining within a PUMA family are more often benign than the substitution matrix alone predicts, and that PUMA genealogy improves molecular function representations compared to treating units independently. A case study of a unit family demonstrates relatedness beyond homology. PUMA achieves competitive performance on downstream tasks when used as a protein language model tokenizer. Moreover, collapsing units into families results in a smaller embedding table and faster training. Together, these results support PUMA as a biologically grounded protein vocabulary that organizes protein units into plausible families of mutational variants. The source code is available at https://github.com/boun-tabi-lifelu/PUMA.
Oct 27, 2024

ProtScape: A molecular structure and energy-aware representation for protein conformation generation

Molecular dynamics (MD) simulations are a principled but computationally expensive approach for studying protein conformational variability, making it challenging to generate large ensembles of structures or characterize transitions between metastable conformations. AI methods for upsampling MD simulations have been developed recently but struggle due to difficulties of sampling complex, high-dimensional molecular distributions. One avenue these methods overlook is to learn a latent space where such sampling becomes easier. To address this, we introduce ProtScape, a generative geometric deep learning framework that learns structure and energy-aware representations of protein conformational landscapes from MD simulations. ProtScape represents protein conformations using an equivariant graph neural network and a multiscale deep wavelet transform that captures local geometric interactions and nonlocal collective motions. This latent representation is dual-organized: 1) by structure captured by wavelet transform layers, via reconstruction error; 2) by energy using a Laplacian energy smoothness penalty. This structured latent space enables meaningful generation and exploration of protein conformations. ProtScape supports multiple generative modes: 1) ensemble generation, which upsamples ensembles given limited MD trajectories flow matching from noise to the organized manifold of conformations, 2) minimum-energy path generation between two high-energy conformations, guided by energy using a nudged elastic band, and 3) energy-descent, which generates trajectories toward lower energies from a high-energy conformation using gradient descent, thereby hypothesizing folding or other stabilizing trajectories. Organizing a latent space by structure and energy lets one representation support ensemble sampling, minimum-energy path finding and energy-guided descent on the same conformational landscape.
Oct 2, 2022

Tensor-reduced atomic density representations

Density based representations of atomic environments that are invariant under Euclidean symmetries have become a widely used tool in the machine learning of interatomic potentials, broader data-driven atomistic modelling and the visualisation and analysis of materials datasets.The standard mechanism used to incorporate chemical element information is to create separate densities for each element and form tensor products between them. This leads to a steep scaling in the size of the representation as the number of elements increases. Graph neural networks, which do not explicitly use density representations, escape this scaling by mapping the chemical element information into a fixed dimensional space in a learnable way. We recast this approach as tensor factorisation by exploiting the tensor structure of standard neighbour density based descriptors. In doing so, we form compact tensor-reduced representations whose size does not depend on the number of chemical elements, but remain systematically convergeable and are therefore applicable to a wide range of data analysis and regression tasks.