benchmarking

How AlphaFold 3 helps with multiscale drug discovery

This study benchmarks AlphaFold 3's interaction scores against BANDOCK+ for drug discovery, showing AF3's slightly better performance in disease ranking but BANDOCK+'s broader coverage.

A prediction-interval framework for identifying outlier control cohorts in lifespan studies

The paper develops a prediction-interval framework to identify outlier control cohorts in lifespan studies using data from the National Institute on Aging Interventions Testing Program.

Do protein representations encode variant function? Diagnosing residue-level protein attributions via embedding-space gradients

This paper evaluates residue-level attributions in protein language models using synthetic fitness landscapes and identifies interventions that improve attribution accuracy.

MutaStruct: Revisiting Benchmarks for Single-Site Mutation-Induced Structural Change Prediction

The paper introduces MutaStruct, a curated benchmark for evaluating mutation-induced structural changes, showing that overall structural fidelity does not guarantee accurate modeling of mutation effects.

SpacePath enables pathway and molecular-family analysis beyond annotation ambiguity in spatial metabolomics and lipidomics

SpacePath improves spatial metabolomics analysis by handling molecular ambiguity and identifying novel lipid families in spatial datasets.

Small-molecule ACOD1 inhibition attenuates colitis and tumor growth through a shared myeloid mechanism, diverging from genetic deletion

This study shows that ACOD1 inhibition reduces colitis and tumor growth via a shared myeloid mechanism, differing from genetic deletion effects.

ResponseVerdict overturns computationally validated immunotherapy signatures, recovers their missed responders and outperforms every published response predictor

ResponseVerdict improves immune checkpoint blockade response prediction by outperforming existing signatures and predictors using data from prior studies.

Vertebrates do not display consistent declines in abundance or stability with contemporary range or realized climate niche centrality

This study finds no consistent decline in vertebrate abundance or stability with distance from range or climate niche centres, challenging assumptions in ecological models.

Anatomically Programmable and Optically Tunable 3D-Bioprinted Tumor-Mimicking Phantoms for Quantitative Nanofluorophore Benchmarking

This paper develops 3D-bioprinted tumor-mimicking phantoms to benchmark nanofluorophores under surgical conditions, showing their ability to evaluate fluorescence properties and guide tumor resection.

Decoding Cell-level Circadian Phase and Rhythm-state Remodeling in Single-cell Transcriptomes With scRhythmBayes

scRhythmBayes infers cell-level circadian phase and rhythm-state remodeling in single-cell transcriptomes, revealing distinct temporal patterns in liver zonation and heart disease.

Optimizing Cognitive Phenotyping for Genetic Discovery in Alzheimer's Disease: A Simulation-Based Evaluation of Mini-Mental State Examination-Based Decline Measures

The paper evaluates cognitive decline measures for genetic studies in Alzheimer's disease using simulations, finding that simpler measures like Slope and dMMSE/dt are more powerful.

Benchmarking Molecular Methods of Protein Engineering and Surveying Protein Search Space to Bolster Artificial Intelligence-Based Designs

The paper benchmarks molecular methods for protein engineering and evaluates AI-based designs by comparing mutagenesis and phenotypic screening of a fluorophore.

SPATNIC enables H&E-grounded cancer-cell identification across spatial and single-cell transcriptomic modalities in colorectal cancer

SPATNIC is a variational autoencoder that identifies cancer cells from spatial and single-cell transcriptomic data in colorectal cancer, outperforming existing tools and identifying a rare cancer stem cell population.

TRUEDA overturns replicated, decontaminated low-biomass disease-microbe associations and clinical decisions, recovers true pathogens and outperforms every decontamination method

TRUEDA identifies true pathogens in low-biomass disease-microbe associations by accounting for contamination, outperforming existing decontamination methods and correcting false associations.

Contribution of the Late Positive Potential to Emotional Memory: A Joint Neuro-Cognitive Approach to Memory Prediction

The paper investigates how the late positive potential (LPP) relates to memory for emotional items, finding that early LPP predicts memory for emotional items but not neutral ones.

Breaking rules at small scales: ecological and biomechanical diversity in ant terrestrial locomotion

This paper synthesizes research on ant locomotion, showing that body size explains much of locomotor performance variation but other factors also contribute to diversity in movement strategies.

Beyond Better Models: Barriers and a Path Forward in Generalisable TCR Antigen Specificity Prediction

This paper identifies barriers to generalisable TCR:pMHC prediction and proposes solutions through data analysis, evaluation setup, and metric alignment.

Withdrawal latency in a human hot plate analog tracks inter-individual differences in pain sensitivity

This study develops a human analog of the hot plate test to measure pain sensitivity through withdrawal latency and vigor, finding that latency correlates with inter-individual pain sensitivity differences.

TACTIC: hybrid denoising and clustering for robust 16S amplicon diversity estimates

The paper introduces TACTIC, a tool that combines denoising and clustering to improve 16S rRNA diversity estimates by reducing inflation caused by within-species sequence variation.

DEEP: Extracting Latent Energetic Landscapes from Protein Structure Predictors

The paper introduces DEEP-Boltz, a framework that extracts energetic potentials from protein structure predictors, outperforming existing methods in predicting mutation-induced stability changes.

Frequency-Resolved Connectivity Graphs for EEG-Based Classification of Alzheimer and Frontotemporal Dementia

This paper develops an EEG-based graph model using frequency-resolved connectivity features to classify Alzheimer's and frontotemporal dementia against healthy controls, achieving competitive accuracy.

Combining RNA-seq and proteomics boosts sensitivity to detect differentially expressed genes by rpMeta

rpMeta integrates RNA-seq and proteomics data to improve sensitivity in detecting differentially expressed genes, identifying new genes and processes in lung cancer subtypes.

Evaluation of object detection neural networks to find topologically associated domains in bulk Hi-C data

This study evaluates YOLOv8 and Mask R-CNN for detecting TADs in bulk Hi-C data by converting Hi-C matrices into images and comparing their performance against reference TADs.

Tracking claim changes from preprint to publication across 72,644 biomedical studies using large language models

In 72,644 matched bioRxiv preprint-publication pairs, LLM-extracted primary abstract claims were unchanged or minorly revised in about 90%; hedging shifted toward caution in 8.4% versus confidence in 4.2%.

When do hybrid models help? Combining mechanistic modelling and machine learning to predict gene regulatory network dynamics

This study investigates when combining mechanistic and machine learning models improves predictions of gene regulatory network dynamics using synthetic data.

Robust Identification of Drug Mechanism of Action Associated Genes Across Heterogeneous Drug Treated Datasets

This paper introduces ConcuRx, a statistical framework that improves gene prioritization across heterogeneous CRISPR drug-response datasets by modeling between-study variability, leading to better identification of MOA-associated genes.

A Community-Driven Single-Cell PBMC Reference Integrating Landmark Datasets Spanning Health and Disease

HARP integrates PBMC data from multiple studies to create a reference atlas with consensus annotations and identifies age, sex, and disease-related cellular modules.

Africa's Genetic Heritage in the Digital Era:DSI Governance Gaps, Multi-Forum Benefit-Sharing Architecture, and Strategic Priorities for COP17.

This paper analyzes Africa's non-human biodiversity and DSI governance gaps, proposing reforms for equitable benefit-sharing and multi-forum coordination.

CascadeMAP: Autonomous Closed-loop Optimization of Enzyme Cascades via Microfluidics, Machine Learning and Agentic AI

Unattended droplet-microfluidic Bayesian optimization of a glycerol detection cascade matched design of experiments (DoE) with one third of the conditions; a second cascade used DoE with Raman readout.

The ontogeny of repeatability: measurement choices shape individuality estimates in a clonal fish

This study shows how measurement practices affect repeatability estimates of individuality in clonal fish, highlighting methodological influences over biological variation.