Discover, rate, and discuss biology preprints.

LipidLibrarian integrates lipidomics data from multiple databases to provide a unified search interface for lipid information.

The paper introduces parallel algorithms to improve the computational efficiency of structured coalescent models for phylogeographic analysis, enabling faster and larger-scale studies of pathogen spread.

scPert predicts single-cell transcriptomic responses to genetic perturbations by integrating large language model embeddings and biological knowledge, improving accuracy in both single-gene and combinatorial perturbations.

TcrDesign generates experimentally validated, functional, full-length TCRs using deep learning, achieving state-of-the-art performance in binding prediction and sequence generation.

The paper introduces MTClass, a machine learning framework for eQTL identification that outperforms traditional methods by detecting more functionally relevant variants in the MHC region.

The paper identifies two gene interaction mechanisms linking ageing and ageing-related diseases using network analysis and machine learning.

This paper introduces a deep-learning model, PALM, that predicts peptide aggregation using protein language model embeddings, showing improved performance with larger datasets.

The paper introduces a framework to infer de novo exonic variants without parental genotypes, enhancing autism association studies by improving gene-based tests and controlling false discovery rates.

This study evaluates using ProtT5 embeddings with gradient boosting for Gene Ontology prediction, finding gradient boosting provides the best balance between performance and stability.

TENGU improves spatial transcriptomics by using transcript-signal density for cell segmentation and simulation, enhancing accuracy in complex tissues.

The paper uses deep learning dynamics to uncover epoch-specific genetic architectures of brain morphology from UK Biobank data.

ActSeekN improves enzyme function annotation by using structural motifs from AlphaFold and UniProt data, outperforming existing methods in yeast, human, and Trichoderma reesei proteomes.

The paper shows that ESM-2 attention maps encode the directional hierarchy of allosteric communication in the adenosine A2A receptor through topological asymmetry.

vcfsim simulates all-sites VCFs with missing data, enabling accurate population genetic studies and benchmarking.

The paper identifies SPP1 and CD44 as influencers of inflammation in preeclampsia by analyzing single-cell transcriptional profiles of placental immune cells.

This study uses single-cell transcriptomics to compare GLP1R expression in human and mouse cardiac tissues and organoids, finding cardiac expression is low in mice but increased in human heart disease.

The paper uses AlphaFold-Multimer to model nAChR subunit stoichiometry in Lepeophtheirus salmonis, challenging assumptions about concatemer design.

GeneLift recovers prognostic gene signatures from underpowered transcriptomic datasets by augmenting data with Gaussian mixture modeling and validating results with EvidenceClock.

STARCall processes in situ sequencing data by integrating image stitching, alignment, and read calling to improve genotype-to-phenotype mapping with reduced error and increased genotyped cells.

MuSIC predicts cross-species RBP-RNA interactions using evolutionary conservation and label smoothing, outperforming existing methods and quantifying variant effects on binding.