Discover, rate, and discuss biology papers.

VINTER is a generative vision-language model that annotates spatial tissues by integrating histology, gene expression, and spatial neighbourhood, outperforming existing methods and transferring across datasets.

PaleoRigor uses a language model with constraints to process paleomicrobiome data, ensuring reproducibility and preventing invalid operations.

The paper shows that additive-input models fail to generalize in Perturb-seq screens when predicting new perturbation targets.

This study analyzes GitHub repositories to show that continuous integration and testing improve software quality and academic impact in scientific software.

The paper introduces fuzzyfold, a Rust-based software framework for stochastic RNA folding kinetics, enabling efficient simulation of long RNA molecules and flexible model integration.

The paper analyzes when pathway scores preserve mediated effects in high-dimensional mediation, introducing a minimax adequacy loss and evaluating performance on TCGA lung adenocarcinoma data.

KENT enables fast metagenomic pathogen screening on low-memory hardware using NVIDIA Jetson Nano clusters, achieving speedups while maintaining accuracy comparable to desktop-based methods.

This paper introduces a semantic precision score to filter generic concepts from biomedical knowledge graphs, outperforming existing methods in drug repurposing queries.

CRAIC provides tools to inspect and curate multiple sequence alignment uncertainty using posterior probabilities from pair-HMM models.

tangermeme provides tools to interpret deep learning models for cis-regulatory analysis, extracting human-readable insights from genomic predictions.

This study identifies and mitigates protocol-dependent gene expression bias between 3' and 5' scRNA-seq by focusing on a small set of biased genes rather than applying broad normalization.

The paper uses spatial and single-cell transcriptomics with mitochondrial lineage tracing to uncover fate switching and premetastatic niche formation in osteosarcoma.

The paper introduces RNA3D-Decoy, a benchmark dataset of AI-generated RNA structure models, to evaluate quality assessment methods for RNA structure prediction.

scStable evaluates scRNA-seq analysis stability by generating synthetic samples with between-sample variation, showing better performance than existing methods in gastric cancer data.

The paper introduces mzQC, a standard for representing quality control metrics in mass spectrometry, enabling interoperable and machine-readable QC reporting.

This study evaluates cross-cohort generalization of mammography diagnostic models using different image representations and domain-adversarial learning, finding that MedSigLIP embeddings provide the best transferability.

CryoCodex improves cryo-EM map quality by jointly enhancing density, estimating local quality, and masking molecules, outperforming existing methods on experimental maps.

FilAdh is a framework for analyzing focal adhesions and actin stress fibers, achieving high accuracy in segmentation, reconstruction, and classification while revealing measurement biases under degradations.

HyperCom infers cell-resolved cell-cell communication using hypergraphs from single-cell transcriptomics data, providing a continuous sender-receiver score for ligand-receptor pairs.

D-plus improves introgression detection by addressing rate heterogeneity and ghost lineages, re-analyzing Jaltomata data to challenge prior interpretations.