Discover, rate, and discuss biology preprints.

The paper develops a random forest model to predict differential protein-protein interactions between experimental conditions using MINT embeddings and mass spectrometry data.

DeepPhage identifies Eu3+-binding peptides from NGS phage display data by analyzing sequence biases and comparing multiple experimental conditions.

This paper introduces a contrastive graph neural network to discover novel larvicidal compounds against Aedes aegypti, identifying LC-79 as a potent and selective candidate.

The paper demonstrates that a well-designed prompting system outperforms model upgrades in generating scientific hypotheses, introducing a framework to create falsifiable, cross-domain hypotheses.

The paper develops deep learning models to predict transcription initiation patterns from DNA sequences across human tissues and cancers.

This study reanalyzes longitudinal serum proteomics data from tuberculosis patients to assess how nutritional-metabolic recovery during treatment varies by baseline BMI, finding no significant BMI-dependent trajectories but an overall increase in a 36-protein summary axis.

LPT uses bounded agentic loops with LLMs and a curated database to design protein binders from natural language prompts, achieving high success rates with low computational costs.

CryoPilot uses two language model agents to automate cryo-EM structure determination, improving resolution and model accuracy through adaptive decision-making.

GaussF improves isoform quantification by using a theoretical Gaussian model to correct sequencing biases without empirical data, outperforming existing tools.

The paper introduces ImmuneTwins, a digital twin framework that personalizes immune response models for individuals, improving prediction accuracy of immune trajectory changes.

This paper introduces multivariate mutual information (MMI) for selecting histone PTM features to predict epigenetic data, showing improved accuracy over traditional methods.

This paper benchmarks active learning strategies for ADMET prediction, finding benefits vary by endpoint and strategy, with Deep Ensembles showing better predictive performance than MC Dropout.

GLESKit is a gene-centered system that synthesizes biomedical evidence by integrating multiple data types, achieving high report quality and performance in benchmark tasks.

SYMBIO-SCAN identifies high-quality full-length 16S rRNA gene sequences from long reads to discover bacterial symbionts in beetles, revealing novel genera and species.

The paper evaluates whether insulation-based losses improve sequence-based prediction of 3D genome folding using a reverse-complement-equivariant state-space model, finding improvements in boundary and loop prediction.

flashscenic accelerates single-cell GRN analysis using GPU acceleration, improving speed and enabling benchmarking of integration strategies in AML data.

iGEX provides an interactive platform for exploring transcriptomic and proteomic datasets, enabling comparison of expression patterns across experiments and public resources.

This paper evaluates H5Nx influenza virus genotyping tools and finds they often group viruses with distinct evolutionary histories due to reassortment and reference bias.

The paper benchmarks GNNs, pretrained models, and fingerprints for ADMET prediction under fair evaluation, finding no single model dominates and highlighting the importance of endpoint-specific selection and leakage-controlled testing.

BaitBench is a tool for designing, assessing, and simulating hybrid capture probes, validated with high correlation to real data.