deep learning

pLM-bVAE: protein language model-enhanced binary representation improves quantum annealing-based protein sequence design

This paper introduces pLM-bVAE, a binary variational autoencoder enhanced with protein language model representations, showing improved sequence reconstruction and design performance in quantum annealing-based protein sequence design.

Multi-scale biophysical modeling of fungal colony growth

This paper develops a multiscale model of fungal colony growth combining discrete hyphal mechanics with a stochastic traveling-wave surrogate to simulate growth efficiently and reveal macroscopic patterns.

Growing a Root from Its Genome : Emergent Tissue Layers and Zones in a GPU-Native Simulation of the Arabidopsis Root Tip

This paper simulates Arabidopsis root tip development from a genome, showing emergent cell identities and zones without pre-defined structures, but with limitations in niche maintenance and regeneration.

SOPSim: A cell vertex-based application for simulating sensory bristle patterning with dynamic filopodia

This study introduces SOPSim, a simulation tool that models sensory bristle patterning in fruit flies by incorporating filopodia dynamics and Notch signaling.

Generative 3D Spatial Reconstruction of Cellular Organization from Single Cell Expression

ORBIS reconstructs 3D tissue geometry from single-cell gene expression without spatial references, achieving lower shape error and faster sampling than existing methods.

Somatic Structural Variant Discovery from Tumor Specific Strings Computed Against a Draft Composite Germline Reference

VESPA-SS detects somatic structural variants from long-read sequencing data without requiring read mapping or a high-quality germline reference, achieving high F1 scores on cancer cell lines.

scCont: Interpretable Contrastive Learning for Functional Characterization of Cellular State Transitions

scCont is an unsupervised contrastive learning framework that extracts interpretable functional transition units from scRNA-seq data, outperforming baseline methods in capturing biologically meaningful dynamics.

Structural Variant Aware Identity-by-Descent Detection on Pangenome Graphs

GIBD detects IBD segments in pangenome graphs and annotates them with structural variants, outperforming existing methods in accuracy and speed.

Virtual dynamics of spatiotemporal single-cell interactions

CytoBridge reconstructs continuous virtual cell dynamics from sparse spatial transcriptomic data across development, regeneration, and neurodegeneration.

Somatic structural variation discovery on cancer pangenome graphs

JANUS discovers somatic structural variants in cancer genomes using a pangenome graph approach, achieving high accuracy and efficiency compared to traditional methods.

Toward space-optimal approximate colored indexes at scale

This paper introduces Zindex, a space-efficient indexing framework for large-scale sequence repositories, achieving compact storage and fast query performance with novel interleaved filters and merging mechanisms.

spearfish: ultrafast Δ-mapping of long reads to redundant transcriptomes

spearfish improves long-read mapping efficiency by leveraging transcriptome redundancy, achieving faster speeds and lower memory usage than existing tools.

Inferring gene regulatory networks from single-cell perturbation data by optimizing causal reach of latent mediator dynamics

CausalODE infers gene regulatory networks from single-cell perturbation data by optimizing causal reach, capturing transient dynamics, and modeling latent protein levels, outperforming existing methods on real datasets.

RPFace: From biased structural benchmarks to generalizable RNA-protein interface prediction

RPFace predicts RNA-protein interfaces directly from sequence, outperforming existing methods on RNA interfaces and generalizing to in vivo data without pretraining.

scCircaSim: A Circadian Simulator for Single-Cell and Spatial Transcriptomics

scCircaSim generates circadian single-cell and spatial transcriptomic data with known ground truth to evaluate analysis methods.

RNA-targeted aminopyridine compounds provide a new antifungal strategy

The paper identifies aminopyridine compounds that target fungal group I introns, showing antifungal activity and synergy with amphotericin B.

Circuit and cell-type mechanisms of the spectrolaminar motif of cortical oscillations

The paper uses a bio-realistic model of mouse primary visual cortex to show that cortical oscillations follow a spectrolaminar motif, with gamma in superficial layers and alpha/beta in deep layers, driven by specific cell types and synaptic properties.

scDisentangle: sequential disentanglement for single-cell perturbation response prediction

scDisentangle predicts single-cell transcriptional responses to perturbations by decomposing expression into residual components, outperforming existing methods in cytokine stimulation, infection, and genetic perturbations.

Multicellular Perturb-seq Reveals Lineage-Specific Transcriptional Control During Human Cardiac Organoid Development

This study uses Perturb-seq with CRISPRi to analyze how cardiac and vascular transcription factors coordinate lineage commitment and cross-lineage signaling in human cardiac organoids.

Rapid, accurate, and memory-frugal sub-gene-level quantification of probe-based, multiplexed single-cell data

This paper introduces simpleaf, a tool for efficient sub-gene-level quantification of probe-based single-cell data, showing improved speed and memory use compared to existing tools.

Active eye modeling: a tool for estimating retinal image quality across the visual field in everyday life

This paper introduces active eye modeling to estimate retinal image quality across the visual field during everyday activities, finding that near work under dim lighting increases peripheral blur.

Plate design confounds donor variation and model evaluation in single-cell perturbation screens

The paper shows that plate design confounds donor variation in single-cell perturbation screens and demonstrates how improved analysis methods can change interpretations of recent studies.

Bioprocess Modeling Made Easy With Hybrax

Hybrax simplifies hybrid bioprocess modeling by providing a framework that automates model setup and data handling, enabling reproducible and reusable models.

Automahout: an autonomous image analysis system for cell tracking

Automahout autonomously tracks cells in time-lapse images using a self-tuning system that annotates, trains, and refines models without human input, achieving high TRA scores on public datasets.

Stoichiometric Encoding of Amplification Kinetics Enables High-Level Multiplexing in Single-Channel TaqMan Real-Time PCR

This paper introduces a method to increase multiplexing in TaqMan qPCR by encoding target-specific information into amplification curves and decoding them with a transformer model, achieving 8-plex detection with high accuracy.

Towards a foundation model for raw nanopore signal through discrete tokenization

This paper introduces Brieno, a self-supervised model for raw nanopore signals that improves read classification and methylation detection without reference genomes.

Single-Cell Transcriptomics Reveals Tumorigenic Lineages and Interactions Persisting and Evolving From Premalignancy to Cancer in the Stomach

The paper uses single-cell transcriptomics to track molecular changes from premalignancy to cancer in gastric tissue, identifying persistent and evolving tumorigenic lineages and markers like EMB.

Clinical relevance of diffusion tensor imaging-based parameters estimated with deep learning: the effect of model retraining

This study evaluates how retraining deep learning models affects the statistical consistency of diffusion tensor imaging parameters in bipolar disorder, showing significant variability across models.

RadMultiBench: A Reproducible Dual-Task Benchmark Exposing the Limits of Pre-Trained Vision-Language Models Across Thoracic and Breast Imaging

RadMultiBench evaluates pre-trained vision-language models on medical imaging tasks, showing performance gaps in report generation and classification across thoracic and breast imaging.

RoDPAA: A Robust Dissimilarity-Based Principal Angle Analysis Framework for Multi-View Microbiome Data Integration

RoDPAA integrates multi-view microbiome data using dissimilarity-based principal angle analysis, recovering shared structure in IBD cohorts where log-ratio methods failed.