dimensionality reduction

NucleiSky enables cross-scale multimodal registration of microscopy data using nuclei constellations

NucleiSky registers microscopy images across scales and modalities by aligning nuclei constellations, achieving accurate localization with minimal landmarks.

Toward Optimizing Thalamic Deep Brain Stimulation for Cortical Modulation: A Surrogate Brain Approach

This paper develops a surrogate-brain approach to optimize thalamic deep brain stimulation for cortical modulation using inferred effective connectivity and control theory.

Connecting single neuron transcriptomes to the projectome in mouse visual cortex

The paper integrates transcriptomic, morphological, and electrophysiological data to define neuron types in mouse visual cortex and predict their projection targets.

Registration-based 3D Light Sheet Fluorescence Microscopy and 2D histology image fusion tool for pathological specimen

LitSHi automatically aligns 3D Light Sheet Fluorescence Microscopy and 2D histology images to improve multimodal tissue analysis.

Cortex-anchored sensor-space harmonics for event-related EEG

This paper introduces a cortex-anchored sensor-space basis for event-related EEG using forward-projected Laplace-Beltrami eigenmodes, showing improved reconstruction efficiency and reliability compared to spherical harmonics and PCA/ICA bases.

Gaining Biological Insights through Supervised Data Visualization

The paper introduces RF-PHATE, a supervised data visualization method that integrates expert knowledge to enhance interpretation of complex biological data.

Faithful Supervised Dimensionality Reduction for Biomedical Data via Decision Geometry

The paper introduces DG-UMAP, a supervised dimensionality reduction method that improves class structure preservation in biomedical data while maintaining local geometry.

Apparent Anatomical Variability Through Rigid Augmentation Enables Reliable Corpus Callosum Segmentation

This study develops a pipeline to generate synthetic MRI scans with anatomical variation and trains a U-Net model to segment the corpus callosum in marmosets, achieving high accuracy in both healthy and disease contexts.

Hundreds of cardiac MRI traits derived using 3D diffusion autoencoders share a common genetic architecture

This study uses 3D diffusion autoencoders to derive latent cardiac MRI phenotypes, finds they share a genetic basis with cardiometabolic traits, and shows predictive power for disease risk.

Metrics for Distinguishing Biological and Interventional Change in AI Models

The paper introduces two metrics to distinguish biological and interventional changes in longitudinal data using ADNI cognitive measurements.

Inverse reinforcement learning reveals action-oriented value signals in naturalistic decision making

This study uses inverse reinforcement learning to identify value signals in the brain during naturalistic decision making, finding strong associations with dorsal striatum activity.

SeedMeasure: an efficient approach and open-source program to quantify seed size

SeedMeasure quantifies seed size and morphology from images, offering a fast, open-source, and cross-platform tool validated across nine species.

Cognitive and affective neurodevelopment in youth exposed to deprivation and threat

The paper examines how deprivation and threat exposure in childhood affect neurocognitive and neuroaffective development in adolescents, finding that both delay neurocognitive maturation beyond socioeconomic factors.

Glomerular Hyperfiltration, Charge Selectivity, and the Low-Dimensional Structure of Glomerular Transport

This study analyzes glomerular sieving data to identify transport modes and shows that principal components explain most variance in transport behavior.

Fault-tolerant 3D reconstruction from 2D spatial proteomics sections

This paper introduces 3D-Omics-Flow, a method that reconstructs 3D molecular volumes from 2D spatial proteomics sections by repairing damaged sections and interpolating between them.

Sparse Distributed Archetypes Reveal Compressible Network Motifs Underlying Naturalistic Cognition

The paper applies Multisubject Archetypal Analysis to fMRI data to reveal sparse, distributed network motifs that capture cognitive state information across multiple scales.

Structural connectivity shapes a cross-modal gradient of brain flexibility

The paper introduces a new measure, flexibility gradient, to quantify regional contributions to brain flexibility and finds that structural properties correlate with regional flexibility contributions across multiple modalities.

Systematic analysis of homozygous autosomal copy number losses in exomes improves diagnostic yield and uncovers ultra-rare recessive disorders

The paper systematically analyzes homozygous autosomal copy number losses in exomes to improve diagnostic yield and identify ultra-rare recessive disorders.

Compound models and Pearson residuals for single-cell RNA-seq data without UMIs

This paper introduces a compound model for single-cell RNA-seq data without UMIs, using Pearson residuals to improve gene selection and embeddings in Smart-seq2 datasets.

The challenges of single cell transcriptomics on difficult human tissue: the placenta

The paper evaluates single cell transcriptomic methods on challenging placental tissue, finding that sample processing and storage have greater impact on results than technical variations.

Unveiling the Terra Cognita of Sequence Spaces using Cartesian Projection of Asymmetric Distances

CAPASYDIS maps sequence spaces into a standardized coordinate system, revealing distinct spatial regions for Bacteria, Archaea, and Eukaryota.

Alpha oscillations support attentional orienting while beta supports perceptual decision-making.

The paper investigates how alpha and beta oscillations relate to attentional orienting and perceptual decision-making, finding distinct roles for each rhythm in these processes.

Entropy-based integration index for quantifying network integration in resting-state functional MRI

This paper introduces an entropy-based framework to quantify network integration in resting-state fMRI by analyzing the distribution of ICA-derived functional contributions across Yeo atlas networks.

CoLa-VAE: A Cell-Cell Communication-Aware Variational Autoencoder for Representation Learning and Expression Denoising

CoLa-VAE improves denoising and representation learning in single-cell data by integrating cell-cell communication information, enhancing downstream analyses like differential communication detection and spatial deconvolution.

ARCADIA Reveals Spatially Dependent Transcriptional Programs through Integration of scRNA-seq and Spatial Proteomics

ARCADIA integrates scRNA-seq and spatial proteomics to reveal spatially dependent transcriptional programs without cell barcode pairing or feature-level correspondence.

ParTIpy: A Scalable Framework for Archetypal Analysis and Pareto Task Inference

ParTIpy scales archetypal analysis for biological trade-offs and integrates with single-cell workflows, enabling task inference in large datasets.

Stimulus and circuit contributions to the information geometry of neural manifolds

The paper develops a differential geometric framework to analyze how neural manifold geometry is shaped by feedforward and recurrent connectivity, showing feedforward connectivity determines representational geometry.

Volumetric Flow Imaging Microscopy to Enhance Particle Characterization

This paper introduces volumetric flow imaging (VFI) to improve particle characterization in flow imaging microscopy, showing increased focus accuracy and resolution for microorganisms.

An atlas-scale generative model for unified representation learning of bulk RNA-seq data

The paper trains a supervised VAE on bulk RNA-seq data to learn a unified tissue-aware representation, achieving high tissue classification accuracy and demonstrating the impact of dataset scale.

The Single-cell Pediatric Cancer Atlas: Data portal and open-source tools for single-cell transcriptomics of pediatric tumors

The paper introduces a data portal and open-source tools for single-cell transcriptomics of pediatric tumors, providing uniformly processed RNA-seq data and metadata from over 700 samples across 55 cancer types.