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.
The paper introduces Boltz2ESI, an end-to-end framework that leverages a biomolecular foundation model to predict enzyme-substrate interactions through native co-folding, thereby capturing active-site plasticity and outperforming existing methods to accelerate biocatalyst discovery and pathway elucidation.
The paper introduces GraNPA, a method that performs GWAS-like association studies on pangenome variation graphs using only a small number of individuals to identify phenotype-related genomic regions without reference bias or additional population data.
DPCGS is a novel computational framework that integrates GWAS summary statistics with single-cell RNA-sequencing data to achieve high-resolution mapping of genetic risk to specific cell subpopulations and regulatory programs, demonstrating superior accuracy over existing methods and providing new insights into the cellular mechanisms of complex diseases like Alzheimer's and asthma.
Through the execution of three Open Calls and three Public Challenges between 2023 and 2025, the AI4Life initiative supported 22 bioimage analysis projects and revealed that while FAIR deep learning adoption faces significant implementation hurdles, the primary bottleneck for scientific AI in biology lies not in method development but in the availability of data, annotations, and shared infrastructure.
Spectronaut-nf is a scalable Nextflow pipeline that enables efficient, parallelized processing of large-scale DIA proteomics datasets on high-performance computing environments, significantly reducing analysis time while maintaining consistent identification results compared to single-workstation or single-node setups.
This paper presents a fully automated, three-stage geometric-to-neural cascade pipeline that combines subject-adaptive unsupervised candidate generation with two lightweight 3D ResNet classifiers trained on minimal human-in-the-loop labels to achieve high-accuracy detection of cerebral microbleeds in susceptibility-weighted MRI while significantly reducing false positives from vascular and non-vascular mimics.
This manuscript has been withdrawn because the foundational "Deviation Error" metric requires substantial mathematical formalization and theoretical validation before it can be established as a proper scoring rule for assessing machine learning predictions.
Spaceland is a morphology-guided computational framework that reconstructs continuous, high-resolution 3D molecular atlases of whole organs by integrating sparse spatial transcriptomics with serial H&E histology, thereby enabling scalable virtual tissue modeling without the need for exhaustive experimental sampling.
This study compares multiple linear regression (MLR), bias-corrected MLR, and the Klemera-Doubal Method (KDM) for estimating biological age from heart rate variability data in 343 subjects, finding that KDM yields the lowest estimation error while the bias-corrected MLR offers improved accuracy over uncorrected MLR but is limited by its reliance on chronological age.
This paper introduces a discrete inverse rendering framework that formulates biological image analysis as a globally optimal integer programming problem, unifying detection, tracking, and event inference to significantly outperform standard local-decision pipelines, particularly in low-signal imaging scenarios.