variational autoencoder

A multisite LFP network (EN-Withdrawal), found with supervised dCSFA-NMF in fentanyl-withdrawing mice, classifies morphine withdrawal, tracks behavioral severity, and its rise predicts fentanyl self-administration escalation in experienced mice.

Haruka identifies condition-specific and shared spatial domains across tissues and conditions using contrastive learning, demonstrating improved detection of spatially heterogeneous perturbation responses.

The paper introduces MeRN, a graph-guided variational autoencoder for learning single-cell metabolic states from transcriptomes, and identifies metabolic drivers in various biological contexts.

This study uses a deep-learning model to analyze scRNA-seq data and identifies how folate deficiency disrupts metabolic transitions in developing mouse neural ectoderm, leading to morphological defects.

The paper applies variational autoencoders to integrate biomedical data, explores model dynamics, and evaluates methods using synthetic and real datasets.

A fully orthogonal IL-2 receptor platform isolates 36 cytokine receptors in T cells; receptor-specific STAT combinations specify distinct fates, and STAT3-dominant programs give curative antitumor responses.

MOFTy disentangles spatial structures and recovers cross-modal patterns in multi-omics data by separating smooth gradients from fine-scale heterogeneity.

CENO is a long-context generative model for genomic sequence understanding and design, showing improved variant prediction and regulatory sequence generation across species.

This paper introduces proteome velocity, a framework using flow matching to infer continuous proteome trajectories from proteomics data, showing its application in mouse sepsis and clinical COVID-19.

Single-cell RNA and BCR sequencing of infected mouse spleens shows class switching overlaps clonal expansion, GC-seeding clones also form plasmablasts, and SHM proceeds at constant rate.

This paper introduces ReconST, a method for optimal gene panel design in spatial transcriptomics, outperforming existing methods in reconstruction accuracy and spatial pattern preservation.

The paper uses a geometry-aware autoencoder to analyze hippocampal activity in mice, revealing a small-world network structure that enables both local fidelity and global searchability in cognitive maps.

The paper introduces PerturbPair, a platform integrating transcriptomic and optical screens to map macrophage regulatory networks, revealing cross-modal relationships and disease-relevant gene programs.

This paper introduces a framework to extract directional functional interactions between neurons, stimuli, and behavior from two-photon calcium imaging data, revealing distinct sensori-behavioral roles in the mouse auditory cortex.

This paper develops a VAE-based framework to translate mouse brain gene expression to human, showing that mouse disease models predict human Alzheimer's and Parkinson's patterns and identify the best mouse model for AD patients.

CoPhaser models biological cycles in scRNA-seq by decomposing data into periodic and non-periodic sources, revealing insights into cell cycle dynamics and other periodic processes across diverse contexts.

GraphBG detects spatial domains in multi-slice and multi-modal spatial transcriptomics data using spectral graph convolutions and Bayesian modeling, achieving high accuracy and scalability.

GRNFormer infers gene regulatory networks from transcriptomics data using a graph transformer framework, outperforming existing methods and recovering known and novel regulatory interactions.

DeepTracing disentangles lineage and intrinsic cellular variation in single-cell lineage tracing data, outperforming existing methods in clustering, phylogenetic recovery, and tumor-type classification.

UNISON is a scalable framework for single-cell transcriptomics that uses skip-zeros variational inference to efficiently analyze large, sparse datasets while preserving interpretability.

TRIO-AI identifies a transitional macrophage population in liver injury using a hybrid model of temporal graphs, ODEs, and VAEs, revealing key signaling pathways and spatial localization.

This paper introduces a latent variable model to analyze hippocampal activity without external labels, revealing continuity-fragmentation dynamics and non-spatial replay in rodent behavior.

BioLACE improves spatial clustering in transcriptomics by integrating spatial geometry and marker gene priors within a VAE framework, achieving better cell type accuracy and interpretability.

SpliceVI integrates sparse single-cell alternative splicing and gene expression data to better capture biological correlates and uncover splicing-based differences in mouse cells.

This paper proposes a deep learning method to align neural network architectures across species for single-cell data, revealing commonalities and differences in gene expression.

scVital integrates scRNA-seq data across species to identify conserved cell states and reduce batch effects using a variational autoencoder and discriminator.

CONCERT, a niche-aware generative model, predicts spatial perturbation responses more accurately than existing methods across multiple tissues and experimental setups.

This paper introduces Dropout Augmentation and DAZZLE, a GRN inference method that improves performance by addressing dropout in single-cell data through model regularization.

ANTIPODE integrates single-cell data across species to identify homologous cell types and reveal gene regulatory divergence in developing mammalian brains.

This paper develops a gated recurrent variational autoencoder to decode decision-making from neuronal data in monkeys, showing comparable accuracy in multidimensional feature space.