1180 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 ESM3-PPISites, a multimodal deep learning model that leverages the ESM3 architecture to accurately identify protein-protein interaction interfaces by integrating evolutionary and 3D geometric contexts, thereby outperforming existing methods and significantly enhancing the efficiency and accuracy of downstream molecular docking simulations.
LoCoFlow is a novel flow-matching framework that enhances single-cell fate inference and trajectory reconstruction by integrating local point-cloud geometry with conditional fate uncertainty, thereby outperforming existing methods in transport accuracy and perturbation response prediction.
The authors introduce rpMeta, an open-source R package that integrates RNA-seq and proteomics data through meta-analysis to significantly enhance the sensitivity of detecting differentially expressed genes compared to analyzing each data type separately.
This study demonstrates that integrating vascular environmental information into a multimodal Transformer-based model significantly improves the prediction of complex astrocyte migration trajectories while establishing a fundamental limit to predictability defined by the irreducibility of movement direction.
Helix Design Studio is an open-source web application that addresses the challenge of designing antimicrobial peptides by utilizing an interpretable, environmental scoring framework to analyze and optimize peptide sequences on a helical scaffold for specific membrane environments while balancing compatibility across different conditions.
This paper demonstrates the feasibility of using locally deployed, open-weight AI agents to coordinate a CRISPR guide-discovery pipeline for diagnostic development, showing that smaller models can effectively support biological research workflows while mitigating restrictions imposed by frontier models.
This paper introduces ProtTale, a generative protein-to-text model equipped with a reliability estimator (pJay) that learns from an LLM judge to identify when its free-form function descriptions are trustworthy, significantly improving annotation precision across various methods without necessarily increasing raw accuracy.
ConforFlux is an inference-time procedure that enhances the diversity of protein conformational sampling in deep learning models like Boltz-2 by introducing a particle-guided repulsion mechanism on trunk embeddings, significantly improving the discovery of multiple states without requiring model retraining.
This paper introduces RiboGRAM, a deep learning model that discovers a multi-scale, positional grammar in non-coding RNA sequences to enable unsupervised structural inference and regulatory motif analysis across diverse organisms and cell types.
This paper introduces BGC Atlas v2, a massively expanded resource containing over 16 million predicted biosynthetic gene clusters with integrated taxonomic, environmental, and geographic metadata to facilitate large-scale natural product discovery and the study of prokaryotic biosynthetic diversity.