variational autoencoder

msiPL: Non-linear Manifold and Peak Learning of Mass Spectrometry Imaging Data Using Artificial Neural Networks

The paper introduces msiPL, a deep learning model for analyzing MSI data, enabling efficient non-linear visualization and identification of biologically relevant patterns.

Hierarchical autoencoder-based integration improves performance in multi-omics cancer survival models through soft modality selection

This paper introduces a hierarchical autoencoder for multi-omics cancer survival analysis, showing improved performance through soft modality selection.

Interpretable deep generative models for genomics

This paper introduces an interpretable variational autoencoder for genomics that identifies gene neighborhoods and hubs, revealing correlations between neuronal differentiation and mitochondrial complex co-expression.

Predicting drug polypharmacology from cell morphology readouts using variational autoencoder latent space arithmetic

This paper uses VAEs to predict drug polypharmacology from cell morphology data and gene expression, finding that beta-VAE and MMD-VAE better disentangle morphology signals.

scMM: Mixture-of-experts multimodal deep generative model for single-cell multiomics data analysis.

scMM is a deep generative model that extracts interpretable joint representations and enables cross-modal generation for single-cell multiomics data, improving clustering and data integration.

Autoencoder networks extract latent variables and encode these variables in their connectomes

The paper uses autoencoder networks to show that regularization can resolve ambiguity in connectivity weights, allowing latent variable extraction through nonlinear dimensionality reduction.

Out-of-distribution prediction with disentangled representations for single-cell RNA sequencing data

The paper uses β-VAE to learn disentangled representations for scRNA-seq data and shows improved out-of-distribution prediction compared to existing methods.

StrainFLAIR: Strain-level profiling of metagenomic samples using variation graphs

StrainFLAIR uses variation graphs to estimate strain-level abundances in metagenomic samples, showing effectiveness on both simulated and real datasets.

Bi-modal Variational Autoencoders for Metabolite Identification Using Tandem Mass Spectrometry

The paper introduces a bi-modal variational autoencoder to identify metabolites from tandem mass spectrometry data by learning from unpaired molecule structures and spectra.

Representation Learning of Resting State fMRI with Variational Autoencoder

This paper uses a variational autoencoder to disentangle sources of variation in resting state fMRI data and shows that latent representations can reliably identify subjects from a large group.

Drop, Swap, and Generate: A Self-Supervised Approach for Generating Neural Activity

This paper introduces Swap-VAE, an unsupervised method for learning disentangled neural activity representations using self-supervised learning on neural recordings.

NSUN2-mediated m5C methylation of IRF3 mRNA negatively regulates type I interferon responses

This study shows NSUN2-mediated m5C methylation of IRF3 mRNA reduces type I interferon responses by promoting IRF3 mRNA degradation, and NSUN2 levels decrease during viral infection to enhance antiviral immunity.

Three-dimensional convolutional autoencoder extracts features of structural brain images with a diagnostic label-free approach: Application to schizophrenia datasets

This paper applies a 3D convolutional autoencoder to structural brain MRI data from schizophrenia and healthy subjects to extract disease-related features without diagnostic labels.

CRISPR-VAE: A Method for Explaining CRISPR/Cas12a Predictions, and an Efficiency-aware gRNA Sequence Generator

This paper introduces CRISPR-VAE, a framework that explains gRNA efficiency predictions and generates efficient gRNA sequences for CRISPR-Cas12a.

Idiopathic chronic diarrhea in rhesus macaques is not associated with enteric viral infections

This study investigates the association between enteric viral infections and idiopathic chronic diarrhea in rhesus macaques, finding no strong link between detected viruses and the condition.

Non-parametric modelling of temporal and spatial counts data from RNA-seq experiments

The paper introduces GPcounts, a non-parametric method for modeling temporal and spatial counts data in RNA-seq using a negative binomial likelihood, showing improved performance over Gaussian and Poisson models.

A universal probabilistic spike count model reveals ongoing modulation of neural variability

This paper introduces a probabilistic spike count model to capture neural variability across time and covariates, finding that variability is modulated by external factors and explained by latent factors.

Extracting Phylogenetic Information of Human Mitochondrial DNA by Linear Autoencoder

The paper uses a linear autoencoder to analyze human mitochondrial DNA sequences, revealing clustering patterns that align with mtDNA haplogroups during training.

Unsupervised Spatial Embedded Deep Representation of Spatial Transcriptomics

SEDR couples a deep autoencoder over expression with a variational graph autoencoder over position, improving clustering in dorsolateral prefrontal cortex and splitting a visually uniform breast tumour region.

MorphVAE: Generating Neural Morphologies from 3D-Walks using a Variational Autoencoder with Spherical Latent Space

The paper introduces MorphVAE, a variational autoencoder for generating neural morphologies from 3D-walks, showing improved cell type discrimination and efficient morphology generation.

Scalable Bayesian GPFA with automatic relevance determination and discrete noise models

This paper introduces a scalable Bayesian Gaussian process factor analysis (bGPFA) for neural data, showing it captures preparatory dynamics and long-timescale processes in primate motor cortex.

Reconstructing unobserved cellular states from paired single-cell lineage tracing and transcriptomics data

The paper introduces TreeVAE, a probabilistic framework to reconstruct unobserved ancestral cellular states from single-cell lineage tracing and transcriptomics data, showing improved performance over benchmarks.

Identifying complex motifs in massive omics data with a variable-convolutional layer in deep neural network

The paper introduces a variable-convolutional layer for motif identification in omics data, showing improved performance over standard convolutional layers.

Latent representation of the human pan-celltype epigenome through a deep recurrent neural network

This paper introduces Epi-LSTM, a deep LSTM autoencoder that creates a pan-celltype epigenomic representation, outperforming existing methods in most cell types and providing smoother genomic representations.

Robust Prediction of Patient-Specific Clinical Response to Unseen Drugs From in vitro Screens Using Context-aware Deconfounding Autoencoder

The paper introduces CODE-AE, a context-aware deconfounding autoencoder, to predict patient-specific drug responses from in vitro screens and discovers personalized therapies and biomarkers.

Sampling from Disentangled Representations of Single-Cell Data Using Generative Adversarial Networks

The paper develops MichiGAN, a generative model combining VAE and GAN strengths, to sample from disentangled single-cell RNA-seq representations and predict drug responses.

Making many out of few: deep generative models for single-cell RNA-sequencing data

This study evaluates deep generative models for scRNA-seq data, finding that they can support experimental design but have limitations in capturing cell type proportions and bimodal gene distributions.

VEHiCLE: a Variationally Encoded Hi-C Loss Enhancement algorithm

VEHiCLE enhances Hi-C contact map resolution using a variational autoencoder and adversarial training, improving downstream analysis capabilities.

ME-VAE: Multi-Encoder Variational AutoEncoder for Controlling Multiple Transformational Features in Single Cell Image Analysis

The paper introduces ME-VAE, a multi-encoder VAE for single-cell image analysis that extracts transform-invariant biologically meaningful features, improving cell population separation and phenotypic differences.

RVAgene: Generative modeling of gene expression time series data

RVAgene is a recurrent variational autoencoder that models gene expression time series data, accurately reconstructing temporal profiles and enabling biological feature discovery.