matrix factorization

Comparison of sparse biclustering algorithms for gene expression datasets

This study compares sparse biclustering algorithms on simulated and real datasets, finding Bayesian algorithms with sparsity constraints perform well on simulated data but are slow, while post-processed non-negative matrix factorization algorithms show strong performance on real datasets.

Agonal factors distort gene-expression patterns in human postmortem brains

This study identifies fever and infection as major agonal factors affecting gene expression in postmortem brains and recommends correcting for these factors using PEER or SVA.

Fast, scalable, and statistically robust cell extraction from large-scale neural calcium imaging datasets

EXTRACT, a GPU-based cell extraction method using one-sided Huber robust regression, processed 199 Allen Brain Observatory sessions (~12 TB) in a day and gave cleaner traces than CAIMAN or CNMF-E.

EMBED: Essential Microbiome Dynamics, a dimensionality reduction approach for longitudinal microbiome studies

EMBED is a dimensionality reduction method for microbiome studies that identifies ecological normal modes to capture community and individual taxa dynamics across longitudinal data.

Binary Interactome Models of Inner- Versus Outer-Complexome Organization

The paper constructs a high-quality yeast binary interactome map and compares inner- and outer-complexome organization, finding the outer-complexome contains more interactions and exhibits higher functional heterogeneity.

The dynamic interactive network of long non-coding RNAs and chromatin accessibility facilitates erythroid differentiation

The paper investigates the interaction between long non-coding RNAs and chromatin accessibility during erythroid differentiation, identifying stage-specific regulatory networks and key lncRNAs involved in the process.

Cell type resolved co-expression networks of core clock genes in brain development.

This study identifies co-expression networks of core clock genes in different cell types during brain development using single-cell RNA sequencing data.

ArchR: An integrative and scalable software package for single-cell chromatin accessibility analysis

ArchR, an HDF5-backed R suite for end-to-end scATAC-seq analysis, outpaces SnapATAC and Signac and processes over 1.2 million simulated cells within 8 h on a laptop.

WEDGE: imputation of gene expression values from single-cell RNA-seq datasets using biased matrix decomposition

WEDGE imputes gene expression values in single-cell RNA-seq datasets using biased matrix decomposition, improving clustering and correlation recovery in sparse data.

Inference and analysis of cell-cell communication using CellChat

CellChat infers communication probabilities between cell groups using a mass-action model and 2,021 curated ligand-receptor interactions (multimeric complexes, cofactors), then compares signaling networks across scRNA-seq datasets.

Unsupervised manifold learning of collective behavior

This paper introduces an unsupervised method to detect and classify collective behavior in complex systems using diffusion maps and covariance analysis, applied to fish schooling data.

Mathematical formulation and application of kernel tensor decomposition based unsupervised feature extraction

This paper introduces a kernel-based tensor decomposition method for unsupervised feature extraction and evaluates its performance on synthetic and real data, showing it outperforms or matches traditional tensor decomposition in large p small n scenarios.

Novel analysis of drug-drug interactions reveals the appearance of non-linear dose dependence under the single drug treatment

This study uses tensor decomposition to analyze gene expression data and finds non-linear dose dependence can occur with single drugs, not just drug combinations.

Amplification of the epigenetic (gestational) age acceleration signal

The paper introduces a method to amplify epigenetic age acceleration signals using regression, PCA, and Lasso, showing significant associations with Downs syndrome, Werners syndrome, prenatal smoking, and maternal BMI.

Dynamics of genome architecture and chromatin function during human B cell differentiation and neoplastic transformation

This study examines 3D genome architecture changes during B cell differentiation and in B cell neoplasias using Hi-C and multi-omic data, revealing dynamic compartment shifts and tumor-specific chromatin reorganization.

Analysis of ovarian transcriptomes reveals thousands of novel genes in the insect vector Rhodnius prolixus

The paper improves the genome and transcriptome of Rhodnius prolixus by identifying over 2,000 novel genes and analyzing stage-specific gene expression during oogenesis.

Correlates of Sleep and Arousal via Matrix Methods

The paper presents a matrix-based method to infer arousal state by integrating neural, metabolic, and immune system data.

Optimal tuning of weighted kNN- and diffusion-based methods for denoising single cell genomics data

This paper introduces DEWAKSS, a method for denoising single-cell genomics data by optimally tuning kNN- and diffusion-based approaches to preserve biological variance and improve clustering.

Object-Attribute Biclustering for Elimination of Missing Genotypes in Ischemic Stroke Genome-Wide Data

The paper addresses missing genotypes in ischemic stroke GWAS data by using biclustering to improve machine learning classifier performance.

Human Chr18: "Stakhanovite" Genes, Missing and uPE1 Proteins in Liver Tissue and HepG2 Cells

The paper investigates the expression of missing and uncharacterized proteins on human Chr18 in liver tissue and HepG2 cells using transcriptomic and proteomic methods, revealing discrepancies in expression levels and method dependencies.

easyMF: A Web Platform for Matrix Factorization-based Biological Discovery from Large-scale Transcriptome Data

easyMF is a web platform that uses matrix factorization to analyze large-scale transcriptome data, enabling biological discovery through metagene-based analysis.

MathFeature: Feature Extraction Package for Biological Sequences Based on Mathematical Descriptors

MathFeature provides a package for extracting mathematical descriptors from biological sequences to improve machine learning performance in bioinformatics.

Meta-analysis of transcriptomes of SARS-Cov2 infected human lung epithelial cells identifies transmembrane serine proteases co-expressed with ACE2 and biological processes related to viral entry, immunity, inflammation and cellular stress.

The paper performs a meta-analysis of RNA-sequencing data from human lung epithelial cells to identify genes co-expressed with ACE2 and related biological processes in SARS-CoV-2 infection.

Gene set inference from single-cell sequencing data using a hybrid of matrix factorization and variational autoencoders

The paper introduces a hybrid model combining matrix factorization and variational autoencoders to infer gene sets from scRNA-seq data, correcting batch effects and identifying cell types.

Towards understanding the regulation of histone H1 somatic subtypes with OMICs.

The paper investigates the regulation of histone H1 somatic subtypes using integrated OMICs data, finding that transcriptional control and co-regulation explain variability in H1 complement and subtype dynamics.

Epigenomic features related to microglia are associated with attenuated effect of APOE ε4 on Alzheimer's disease risk in humans

The paper identifies an epigenomic factor linked to reduced activated microglia that mitigates the risk of Alzheimer's disease in APOE ε4 carriers.

Microstructural underpinnings and macroscale functional implications of temporal lobe connectivity gradients

The paper identifies three structural connectivity gradients in the human temporal lobe and links them to microstructural features and macroscale functional measures.

Sparse Project VCF: efficient encoding of population genotype matrices

The paper introduces spVCF, a compressed variant call format that reduces file size by over 10X with minimal information loss.

Real-Time Point Process Filter for Multidimensional Decoding Problems Using Mixture Models

The paper proposes a computationally efficient approximate point process filter using mixture models for real-time decoding of neural spiking activity in multidimensional spaces.

A comprehensive analysis of the global human gut archaeome from a thousand genome catalogue

The paper analyzes 1,167 archaeal genomes from the human gut to reveal novel species, genomic adaptations, and correlations with socio-demographic factors.