Discover, rate, and discuss biology preprints.

The paper presents a Python implementation of scDblFinder with GPU support for faster doublet detection in single-cell RNA sequencing data.

The paper introduces a pipeline for assessing and comparing genome assembly quality using multiple metrics and tools.

CryoForge is an evidence-gated agent that improves cryo-EM model building by selectively accepting or rejecting edits based on structural evidence, achieving validated improvements in most test cases.

This study shows that protein size, geometry, and oligomeric state systematically influence mutational robustness and evolvability.

GlycoMeSH links glycan structures to biomedical contexts through a database and enrichment workflow, improving association recovery and coverage over existing methods.

PLUM generates antimicrobial peptides with controlled function and length, achieving high yields and diverse sequences with favorable predicted properties.

A logistic regression on STRING evidence picked DUF (domain of unknown function) protein pairs for AlphaFold 3; of 12,298 modelled pairs, 1,208 met strict and 2,433 liberal confidence.

The paper identifies 277 putative biosynthetic gene clusters for benzoxazolinate-like molecules through genome mining and presents a neural network model for detecting rare BGCs.

MARLOWE is a computational tool for taxonomic characterization of unknown forensic samples using de novo peptide identification, achieving high specificity in species-level identification.

This study benchmarks docking protocols for GPCR allosteric modulators using experimental structures and GaMD simulations, finding that ensemble re-ranking strategies improve early enrichment of known actives.

PubMind extracts and annotates genetic variant functionality from biomedical literature using LLMs, creating a database with over 1.3 million variants.

BART-spatial identifies transcriptional regulators from spatial omics data by integrating spatial and pseudotemporal information with TR binding profiles, outperforming existing methods.

PandaDock is an open-source molecular docking platform that improves pose prediction and affinity scoring using a GNN and flexible-ligand search, achieving strong performance on benchmark datasets.

The paper presents a machine learning model that predicts pathogenic DNA variants using a semantic approach combined with CRF and BiLSTM, achieving state-of-the-art performance.

scUnify provides a unified framework for training and inference across multiple single-cell foundation models, enabling comparison and extension of tasks and adaptation strategies.

The paper develops a label-free framework to detect antimicrobial resistance determinants using evolutionary signatures and protein-language models, finding known determinants are mostly accessory in E. coli but enriched in K. pneumoniae.

This paper introduces TreeOwen, algorithms for computing Owen values in tree-based ensembles to provide structured, multi-resolution explanations for grouped features in scientific applications.

GEOMeta automates metadata curation for GEO transcriptomic data, generating standardized annotations for 600,000 human bulk RNA-seq samples and benchmarking LLMs for annotation quality.

MSMICA improves metabolite identification in untargeted metabolomics by integrating mass spectrometry data, retention time, and biological evidence, achieving high accuracy across mammalian datasets.

This paper evaluates circulating microbial DNA as a cancer biomarker, finding that observed patterns are largely artifacts of sequencing and analysis methods rather than biological signals.