850 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 improved 3D radial phyllotaxis trajectories, specifically UPhy and FlexiPhy, which achieve uniform readout direction density and utilize randomized interleave ordering to significantly reduce ringing artifacts and enhance the robustness of retrospective sequential binning in dynamic MRI.
SSUplex is a fast, open-source Rust tool that detects, extracts, and origin-sorts full-length small-subunit rRNA reads from both DNA strands into five categories (bacteria, archaea, eukaryota, mitochondria, and chloroplast), offering a high-speed, memory-efficient alternative to Metaxa2 for environmental DNA metabarcoding workflows.
The paper introduces TEDlm, a domain-centric protein language model pretrained on structurally defined domain segments that outperforms larger full-sequence models in remote-homology detection and molecular function prediction, demonstrating that focusing on domain-intrinsic signals yields compact, structurally informed representations.
The amR R package suite offers a comprehensive, interpretable framework for predicting antimicrobial resistance in bacterial pathogens by integrating multi-scale genomic feature extraction, machine learning model training, and interactive visualization to uncover cross-species and multi-drug resistance mechanisms.
This paper evaluates AlphaFold3's performance across diverse biomolecular applications, finding that while it offers powerful all-atom modeling capabilities, its accuracy and reliability are uneven and heavily dependent on training-set overlap, necessitating cautious interpretation compared to its predecessor.
The Genomic Annotation Infrastructure (GAIn) is a platform that enables transparent, reproducible, and scalable genomic variant annotation through declarative pipelines, public resource repositories, and flexible web and command-line interfaces that support custom extensions and automated re-annotation.
This study reveals that in the multimodal protein language model ESM3, distinct physical modalities (sequence, structure, secondary structure, and solvent accessibility) initially occupy separate subspaces before fusing into a shared low-dimensional representation between layers 25 and 35, while functional annotations remain orthogonal throughout, with this fusion process being a learned, universal property independent of protein length but delayed by structural disorder.
ProtBLIP2-SST is a novel two-stage framework that integrates protein sequence and 3D structural information via a structure-aware language model and a Q-Former projector to enable large language models to generate flexible, open-ended functional captions, thereby overcoming the limitations of traditional rigid gene ontology classification.
EcoMorph is a modular system that leverages the prompt-based Segment Anything Model 3 (SAM3) to accurately quantify morphological traits like area, size, and abundance across diverse ecological contexts without requiring taxon-specific training, thereby enabling high-throughput ecological research from heterogeneous image sources.
The paper introduces CellTok, a novel approach that tokenizes continuous single-cell transcriptomic profiles into discrete sequences compatible with pretrained large language models, thereby enabling a unified framework for jointly processing cellular data, biological context, and textual instructions to perform diverse tasks like cell identification, disease inference, and state generation.