variational autoencoder

Integrating single-cell RNA-seq datasets with substantial batch effects

This paper improves scRNA-seq dataset integration by introducing regularization strategies for cVAEs, enhancing batch correction while preserving biological variation.

Calibrated Identification of Feature Dependencies in Single-cell Multiomics

This paper introduces VI-VS, a framework for identifying conditionally dependent features in multiomic data while controlling false discovery rates.

DynaMight: estimating molecular motions with improved reconstruction from cryo-EM images

DynaMight improves cryo-EM reconstruction by estimating molecular motions through 3D deformation learning and error estimation using VAEs.

Modeling functional cell types in spike train data

The paper develops a method to simultaneously fit single-cell neural activity models and group neurons into functional cell types, improving prediction accuracy and interpretability.

Isolating structured salient variations in single-cell transcriptomic data with StrastiveVI

StrastiveVI isolates aging and Alzheimer's disease-associated patterns in scRNA-seq data by separating them from sex, tissue, and cell type variations.

Benchmarking Variational AutoEncoders on cancer transcriptomics data

This study benchmarks variational autoencoders on cancer transcriptomics data to evaluate hyperparameter effects and latent space biological relevance.

Representational Learning from Healthy Multi-Tissue Human RNA-seq Data such that Latent Space Arithmetics Extracts Disease Modules

This study uses a variational autoencoder trained on healthy human RNA-seq data to predict disease gene expression changes and identify drug targets through latent space arithmetic.

PAST: latent feature extraction with a Prior-based self-Attention framework for Spatial Transcriptomics

PAST characterizes spatial domains in spatial transcriptomics data using a variational graph convolutional auto-encoder with prior-based self-attention, improving downstream analyses.

Variational Log-Gaussian Point-Process Methods for Grid Cells

The paper develops variational log-Gaussian point-process methods to efficiently calculate spatial statistics for grid cells in large environments using Gaussian processes.

OPUS-Fold3: a gradient-based protein all-atom folding and docking framework on TensorFlow

OPUS-Fold3 is a gradient-based protein folding and docking framework that generates accurate 3D structures under specified constraints, outperforming pyRosetta in side-chain modeling.

Deep generative modeling of transcriptional dynamics for RNA velocity analysis in single cells

This paper introduces veloVI, a deep generative model for RNA velocity analysis, showing improved data fit, consistency, and uncertainty quantification compared to existing methods.

Divide-and-conquer quantum algorithm for hybrid de novo genome assembly of short and long reads

This paper proposes a hybrid quantum algorithm for genome assembly using short and long reads to address repetitive sequences and sequencing errors, showing improved convergence and optimization.

Representation of Protein Dynamics Disentangled by Time-structure-based Prior

The paper introduces two methods with a time-structure-based prior to disentangle protein dynamics from MD simulations, improving the representation of conformational transitions.

Towards interpretable learned representations for Ecoacoustics using variational auto-encoding

This paper develops a variational auto-encoder to learn interpretable acoustic features from ecoacoustic data and demonstrates its ability to infer habitat degradation levels.

Deep unsupervised learning methods for the identification and characterization of TCR specificity to Sars-Cov-2

This study uses deep unsupervised learning methods to identify and characterize TCR specificity to Sars-Cov-2, achieving promising results in understanding TCR patterns.

In search of disentanglement in tandem mass spectrometry datasets

This study uses variational autoencoders to disentangle factors in tandem mass spectrometry data, enabling better interpretation and comparison across instruments.

Onsager's variational principle in proliferating biological tissues, in presenceof activity and anisotropy

This paper applies Onsager's variational principle to model proliferating biological tissues, considering activity and anisotropy, and derives coupled partial differential equations for growth and chemical interactions.

Switching state-space modeling of neural signal dynamics

This paper develops a switching state-space model for analyzing time-varying neural data, showing improved performance in detecting transient sleep spindles from EEG recordings.

SIMVI reveals intrinsic and spatial-induced states in spatial omics data

SIMVI disentangles intrinsic and spatial-induced variations in spatial omics data, revealing spatial effects and biological dynamics across multiple datasets.

Mechanisms of active diffusion of vesicular stomatitis virus inclusion bodies and cellular early endosomes in the cytoplasm of mammalian cells

This study investigates the random motion of VSV inclusion bodies and early endosomes in mammalian cells, revealing that their movement is mediated by trap movement rather than hopping between fixed traps.

Target-aware Variational Auto-encoders for Ligand Generation with Multimodal Protein Representation Learning

The paper introduces TargetVAE, a variational auto-encoder for ligand generation using multimodal protein representations, showing improved binding affinity and generative quality.

Isolating salient variations of interest in single-cell data with contrastiveVI

The paper introduces contrastiveVI, a method to disentangle shared and treatment-specific variations in scRNA-seq data, showing improved performance over existing methods in four datasets.

Protein Ensemble Generation through VariationalAutoencoder Latent Space Sampling

The paper generates high-quality protein conformations using a variational autoencoder, achieving better sampling accuracy than MD simulations or AlphaFold2 for K-Ras.

Generative Modeling of Single Cell Gene Expression for Dose-Dependent Chemical Perturbations

This paper introduces scVIDR, a variational autoencoder for predicting dose-dependent gene expression in single cells across different cell types and tissues, showing improved performance over existing methods.

Variational Autoencoders for Generating Synthetic Tractography-Based Bundle Templates in a Low-Data Setting

This study uses a variational autoencoder to generate synthetic tractography-based bundle templates from a small dataset, showing improved shape capture compared to standard atlases.

Synthetic whole-slide image tile generation with gene expression profiles infused deep generative models

This paper introduces RNA-GAN, a deep generative model that creates synthetic whole-slide image tiles using gene expression profiles, outperforming traditional GANs and showing imputation capabilities.

Variational autoencoders to predict DNA-methylation age and provide biological insights in age-related health and disease

This paper uses variational autoencoders to predict DNA-methylation age and identifies biologically relevant features of ageing.

How many variables does Wordclim have, truly? Generative A.I. unravels the intrinsic dimension of bioclimatic variables.

The paper uses a variational autoencoder to reduce 19 bioclimatic variables to 5 uncorrelated dimensions, showing they produce better species distribution models.

Biologically informed variational autoencoders allow predictive modeling of genetic and drug induced perturbations

This paper introduces OntoVAE, a VAE that incorporates ontologies to predict genetic and drug-induced perturbations using bulk and single-cell transcriptomic data.

Learning consistent subcellular landmarks to quantify changes in multiplexed protein maps

The paper introduces CAMPA, a deep-learning framework for analyzing subcellular organization changes, and applies it to human cells to study how perturbations affect subnuclear organization and membraneless organelles.