388 papers
Biophysics sits at the fascinating intersection where the laws of physics meet the complexity of living systems. This field uses tools like light, electricity, and mechanical forces to decode how cells move, how proteins fold, and how our senses translate the world around us. Rather than just observing biology, biophysicists measure and model life to understand the fundamental machinery that powers every organism.
On Gist.Science, we make these discoveries accessible by curating the latest preprints directly from bioRxiv. Our team processes every new submission in this category, providing both clear, plain-language overviews and detailed technical summaries so readers of all backgrounds can grasp the cutting-edge science. Below are the most recent biophysics papers from bioRxiv, ready for you to explore.
The paper introduces BioSecBench-Function, a verifiable benchmark comprising 111 evaluations across seven biosecurity-relevant question types to assess the ability of AI agents to accurately infer biological functions from experimental data, revealing significant performance variations among models and highlighting that cost is not a reliable predictor of accuracy.
This study presents a minimal synthetic cell constructed from Giant Unilamellar Vesicles equipped with gated DNA nanopores that autonomously maintain homeostasis by utilizing a negative feedback loop where internally transcribed RNA blockers regulate rNTP influx, enabling sustained RNA production for up to 16 hours under varying external conditions.
This study presents the first genome-wide analysis of human blood transcriptional responses to neutron versus photon irradiation, revealing that neutrons induce a more potent and distinct gene expression profile driven by cGAS-STING-NF-κB signaling with a global transcriptomic relative biological effectiveness (RBE) of approximately 1.3, while identifying a narrow 6-hour time window and specific gene signatures essential for distinguishing neutron exposure from photon exposure in biodosimetry.
By employing an optogenetic toolbox to manipulate actin crosslinker length, this study demonstrates that the mechanical properties of the actin cortex, specifically its stiffness and stress-stiffening behavior, encode a length-dependent switch that determines whether cells undergo cortical delamination or polarization and symmetry breaking.
This paper presents the structure of an RNA polymerase ribozyme replication complex, revealing how directed evolution shaped a conserved catalytic core into a "right hand" configuration similar to protein polymerases to facilitate the copying of genetic material.
This paper presents a thermodynamic framework that utilizes sequence-derived entropy estimates and protein language models to map, classify, and predict elastic proteins across the human proteome, revealing distinct recoil mechanisms and expanding their functional scope beyond traditional structural materials.
This study employs integrated molecular dynamics and machine learning approaches to elucidate the atomic-level mechanisms of ligand recognition and gating in the ancestral insect odorant receptor MhOR5, revealing distinct binding modes, dual release pathways, and specific molecular determinants that govern the differential stability and kinetics of eugenol and DEET interactions.
This study demonstrates that while both the McIntyre-Richardson-Grill (MRG) and Cohen double-cable axon models effectively predict deep brain stimulation outcomes, the more detailed Cohen model exhibits greater excitability, lower activation thresholds, and superior predictive accuracy, particularly regarding frequency-dependent responses.
This study demonstrates that computational structural comparisons of POLG variants R627Q and W748S using AlphaFold2 fail to distinguish mutation-specific effects from model-selection variability, though experimental structure mapping suggests these residues share a local microenvironment where the mutations likely disrupt specific polar and hydrophobic interactions rather than direct macromolecular contacts.
The paper introduces WaterFlow, a flow-matching-based model that outperforms existing state-of-the-art methods in predicting ordered water molecule positions on protein structures with sub-angstrom accuracy, thereby enhancing applications in protein design, drug discovery, and structural refinement while revealing that the diversity of high-quality training data currently limits further performance gains.