dimensionality reduction

scCotag: Diagonal integration of single-cell multi-omics data via prior-informed co-optimal transport and regularized barycentric mapping

scCotag improves single-cell multi-omics integration by using co-optimal transport and regularized barycentric mapping, outperforming existing methods in alignment and embedding accuracy.

Spatial Polarization-Induced Fluorescence Fluctuation Imaging (SPIFFI) Enables Single-shot Super-Resolution and Multidimensional Imaging

SPIFFI enables single-shot super-resolution and six-dimensional imaging by leveraging polarization-induced fluorescence fluctuations, enhancing live-cell imaging capabilities.

TWO STEPS AWAY FROM EXTINCTION: DO YOU BELIEVE?

The paper examines the effects of forest fragmentation on genetic diversity and demographic structure of the critically endangered tree species Dinizia jueirana-facao, finding reduced pollen dispersal distances and no evidence of inbreeding or genetic diversity loss.

Regulatory element modules as universal features for single-cell chromatin analysis

The paper introduces REMO, a set of DNA regulatory element modules for single-cell chromatin analysis, improving cell state separation and annotation.

Global signal regression reduces connectivity patterns related to physiological signals and does not alter EEG-derived connectivity

This study examines the impact of global signal regression on resting-state fMRI connectivity, finding it reduces physiological artifacts while preserving electrophysiological connectivity patterns.

Cognitive Flexibility and Brain Network Energy in Healthy Aging: an Allostatic Perspective from the SENECA model

The paper analyzes resting-state fMRI data to show how brain network energy redistributes with age, supporting cognitive flexibility through an allostatic mechanism.

Single-objective lattice light sheet microscopy with microfluidics for single-molecule super-resolution imaging of mammalian cells

This paper introduces a single-objective lattice light sheet microscopy platform for improved 3D single-molecule super-resolution imaging in scattering samples.

EMCF ecosystem: Towards pretrained foundation model for electron microscopy image analysis

The paper introduces EMCFsys, an ecosystem for electron microscopy image analysis, improving image restoration and analysis efficiency with a benchmark dataset, foundation model, and restoration model.

SEAL: Spatially-resolved Embedding Analysis with Linked Imaging Data

SEAL integrates spatial imaging context with dimensionality reduction embeddings to improve interpretation of complex datasets through interactive visualization and feature importance analysis.

CellMentor: Cell-Type Aware Dimensionality Reduction for Single-cell RNA-Sequencing Data

CellMentor improves dimensionality reduction for scRNA-seq by preserving cell-type-specific signals and reducing batch effects across datasets.

Randomized Spatial PCA (RASP): a computationally efficient method for dimensionality reduction of high-resolution spatial transcriptomics data

RASP is a fast dimensionality reduction method for spatial transcriptomics data that improves computational efficiency while maintaining accuracy in spatial domain detection.

Solving the gene classification problem with a novel Gene Space approach

This paper introduces a Gene Space framework to characterize genome organization by analyzing nucleotide token frequencies, enabling species-specific gene classification and evolutionary pathway constraints.

Morphometric Latent Factors in Autism and Their Association with Receptor Profiles and Behavior

This study compares latent factors from functional connectivity and morphometric measures in ASD, finding that morphometric factors better correlate with behavioral traits and neurodevelopmental mechanisms.

MultiSpecies Canopy Segmentation: Interactive Machine-Learning and Pseudo-Labelling are key

This paper develops a method to train deep-learning models for multi-species canopy segmentation using FCNNs and pseudo-labeling, achieving high accuracy with reduced human effort.

Cell type- and chromosome-specific chromatin landscapes and DNA replication programs of Drosophila testis tumor stem-cell like cells

Neural correlates of kinematic features of passive finger movement revealed by univariate and multivariate fMRI analyses

The paper reanalyzes passive finger movement fMRI data using univariate and multivariate methods to show that multivariate analysis better identifies brain regions encoding kinesthesia and proprioception, with larger neural representation for amplitude and direction than velocity.

Fast Fourier Transform is a training-free, ultrafast, highly efficient, and fully interpretable approach for epigenomic data compression

This paper introduces a training-free FFT-based method for efficient and interpretable epigenomic data compression, achieving high fidelity and speed compared to autoencoders.

A generalized and versatile framework to train and evaluate autoencoders for biological representation learning and beyond: AUTOENCODIX

AUTOENCODIX provides a standardized framework for training and evaluating autoencoders, demonstrating its utility across multi-modal biological data and offering insights into representation learning trade-offs.

Leveraging CryoEM and AI-Driven Morphological Feature Analysis for Insights on Bacterial Structures

This paper develops an AI-driven workflow for automated analysis of bacterial ultrastructure in cryoEM data, enabling rapid and accurate quantification of membrane thickness, flagella length, and bacterial arrangement.

Unified Probabilistic Analysis of CyTOF: A Deep Generative Approach using CytoOne

CytoOne is a probabilistic framework for CyTOF data that integrates batch correction, differential analysis, and visualization, effectively modeling sparse and noisy data.

Phenotype-driven parallel embedding for microbiome multi-omic data integration

This paper introduces PAPRICA, a framework for integrating microbiome multi-omic data by preserving omic-specific properties and phenotype associations, showing improved performance across datasets.

VegVault dataset: linking global paleo-, and neo-vegetation data with functional traits and abiotic drivers

The paper introduces VegVault, a database integrating paleo- and neo-vegetation data with functional traits and abiotic drivers to support ecological research on biodiversity patterns.

Comparative Analysis of Feature Selection Methods for Single-Cell RNA Sequencing Data

The paper compares 16 feature selection methods for scRNA-seq data across three datasets, finding substantial variability and identifying a core set of genes consistently selected by most methods.

IsoSpace: Chemistry-Informed Dimensionality Reduction and Automated Isotope Candidate Detection in Imaging Mass Spectrometry

IsoSpace provides a chemistry-informed dimensionality reduction method for imaging mass spectrometry that identifies isotopic patterns without prior assumptions.

Cortex-Wide Preservation of Multi-Stimulus Information across Degrees of Network Synchronization

The paper decodes stimulus identity and network state from calcium imaging in mice across different anesthesia states, showing preserved stimulus information even during deep anesthesia.

Assessing the robustness of deep learning based brain age prediction models across multiple EEG datasets

This study evaluates deep learning models for predicting brain age from EEG data across multiple datasets, finding that performance varies with dataset shifts and that certain hyperparameters improve generalization.

Plant Face: Machine learning decodes genetic, environmental and developmental imprints in leaf appearance

This paper uses deep learning to show that individual plant leaves are unique, revealing genetic, environmental, and developmental information in their appearance.

Sharpened visual memory representations are reflected in inferotemporal cortex

The paper shows that inferotemporal cortex, like the hippocampus, reflects sharpened visual memory representations, contributing to visual memory fidelity.

Novae: a graph-based foundation model for spatial transcriptomics data

Novae is a graph-based foundation model for spatial transcriptomics that improves domain inference and batch correction across multiple tissues and technologies.

CellUntangler: separating distinct biological signals in single-cell data with deep generative models

CellUntangler disentangles multiple biological signals in single-cell data using deep generative models, separating cell cycle from other signals like cell type and differentiation.