1012 papers
Bioinformatics sits at the exciting intersection where biology meets data science, using powerful computer tools to decode the vast complexity of living systems. From mapping the human genome to tracking how viruses evolve, this field transforms raw biological information into actionable insights that drive modern medicine and research forward without requiring a supercomputer to understand the basics.
On Gist.Science, we ensure you never miss a breakthrough by processing every new preprint in this category directly from bioRxiv. Our team provides both plain-language explanations and detailed technical summaries for each paper, making cutting-edge discoveries accessible to everyone regardless of their background.
Below are the latest bioinformatics papers added from bioRxiv, ready for you to explore with clarity and depth.
This paper introduces nf-sarcopipe, a modular Nextflow pipeline that integrates de novo and reference-guided miRNA discovery with transcriptomic analysis to identify novel and known exercise-responsive miRNAs and their regulatory networks involved in skeletal muscle adaptation.
This study benchmarks three codon-focused masked language models against traditional methods, demonstrating that while no single model dominates all tasks, they effectively capture translational context to generate variants with superior expression performance, suggesting that sampling across multiple models is a promising strategy for optimizing nucleic acid-based medicines.
This paper demonstrates that previously reported advantages of codon language models in synonymous-variant prediction are evaluation artifacts caused by data leakage, as these gains disappear under strict leakage-controlled benchmarks like the newly released CodonBench.
The paper introduces scDRP, a generative framework utilizing disentangled representation learning and conditional optimal transport to accurately estimate individualized treatment effects and infer counterfactual cell states from unmatched single-cell perturbation data, thereby revealing heterogeneous biological mechanisms across diverse cell types and conditions.
The paper presents "HRA: Powers of Ten," an open-source virtual reality application that leverages the Human Reference Atlas to immerse users in a multiscale, semantically harmonized exploration of single-cell biomedical data, enabling new biological insights into cellular architecture and senescence across organs and spatial scales.
The authors present pastForward, a fully automated Snakemake pipeline that streamlines ancient and historical DNA analysis by integrating user-friendly processing, eukaryote-wide taxonomic screening, and copy-number variation tracking to enable longitudinal genomic studies across diverse species.
PARNET is a novel RNA foundation model trained exclusively on experimental CLIP-seq data to predict base-resolution RBP binding profiles, demonstrating superior performance and mechanistic interpretability across diverse downstream tasks compared to traditional sequence-based language models.
The paper introduces TriTower-m6Am, a triple-tower deep learning architecture that integrates semantic, sequential, and structural RNA representations to significantly improve the accuracy and interpretability of m6Am site prediction compared to existing single-representation methods.
This paper introduces ASPIRE, an accessible Nextflow workflow designed to process, analyze, and interpret linked amplicon sequence variant (ASV) and volatile organic compound (VOC) data to investigate respiratory microbial ecosystems and their relationship to health.
The paper introduces megaMine, a transparent and scalable rule-based framework that effectively extracts structured gene-cancer-drug evidence from biomedical literature, demonstrating high accuracy in distinguishing therapeutic efficacy and generating interpretable data for downstream knowledge synthesis.