1769 papers
The Artemis I mission induced genotype-dependent changes in seed amino acid profiles and post-flight growth in *Arabidopsis thaliana*, revealing that branched-chain amino acid metabolism is a critical factor influencing plant responses to deep-space exposure.
This study establishes a robust framework integrating UAV-based multispectral phenotyping and time-series genome-wide association studies to identify dynamic genetic loci and the candidate gene *Zm00001d048615* that regulate maize leaf area index, offering valuable targets for improving canopy architecture and yield.
This study challenges the long-held view that spermiogenesis involves global transcriptional silencing by demonstrating that the haploid genome remains transcriptionally active and dynamically organized throughout this phase, necessitating a redefinition of late spermatid differentiation as a period of transcriptional specialization rather than shutdown.
This study demonstrates that a hybrid machine learning model utilizing daily self-reported symptom variability can accurately classify four menstrual cycle phases without wearable sensors, offering a scalable and equitable solution for global reproductive health monitoring.
This study utilizes bioinformatics analysis and experimental validation to identify DUSP1 and EGR1 as key molecular regulators linking traumatic brain injury to post-traumatic epilepsy, suggesting them as promising therapeutic targets.
This paper introduces DynoM, a generative deep learning framework trained on protein complex interactions that significantly outperforms existing monomer-only models in predicting accurate conformational ensembles and suppressing structural hallucinations by leveraging the long-range physical constraints inherent in complex data.
This study provides field evidence that in low-flow streams, fish eDNA signals exhibit a delayed downstream appearance and persist for a period after the source organism is removed, highlighting the need to account for transport dynamics and temporal persistence when interpreting eDNA data in lotic ecosystems.
This study constructed and validated a robust machine learning-based prognostic model using 32 tumor microenvironment-related genes to predict biochemical recurrence in prostate cancer, revealing distinct immune landscapes and potential therapeutic sensitivities for risk-stratified patient management.
This study demonstrates that microsaccade dynamics serve as a non-volitional physiological biomarker for expertise in chest radiography, as increased microsaccade production during nodule fixation correlates with successful detection and distinguishes higher-accuracy readers from less experienced ones.
This study identified three triaptosis-related prognostic genes (BNIP3L, KLHDC1, and PRDX3) in breast cancer through bioinformatic analysis and RT-qPCR validation, establishing a predictive nomogram that links these genes to oxidative phosphorylation pathways, immune infiltration, and potential therapeutic targets.