3724 papers
By analyzing lung adenocarcinoma and colorectal cancer, this study maps the landscape of functional plasticity within the tumor microenvironment, revealing tumor-specific transitory myeloid states associated with immune suppression and widespread convergence of diverse cell types into similar functional states.
This independent computational study demonstrates that machine learning classifiers, particularly Random Forest, can effectively distinguish CTCF-associated DNA sequences from dinucleotide-shuffled backgrounds using short k-mer frequency features, thereby establishing a transparent and reproducible end-to-end workflow for integrating molecular biology with applied machine learning.
This metazoan-wide phylogenomic study demonstrates that the canonical eight-family structure of prostanoid receptors is a vertebrate-specific innovation, revealing that previously proposed non-vertebrate homologs are actually distinct receptor lineages whose apparent similarity to vertebrate EP4 receptors stems from phylogenetic proximity rather than true orthology.
This study demonstrates that sea buckthorn fruit extract exerts osteoprotective effects in ovariectomized mice by promoting bone formation and suppressing bone resorption through an ERα-mediated upregulation of β-catenin signaling.
Through a large-scale, multitaxon analysis using environmental DNA, this study demonstrates that non-perennial rivers host unique and diverse communities distinct from both perennial rivers and terrestrial soils, challenging the view of them as merely transitional systems and highlighting their critical role in global biodiversity.
This systematic review synthesizes heterogeneous human evidence on the gut microbiome in sickle cell disease to propose an evidence-informed framework that moves beyond inconsistent taxonomic profiles toward testable hypotheses regarding functional metabolic alterations.
By applying a novel simulated annealing framework to single-cell RNA-seq data from four human cancers, this study identifies five conserved T cell exhaustion drivers and twenty novel candidate genes while revealing tissue-specific exhaustion signatures that challenge the universal role of previously reported markers like NR4A1.
This study reveals that invasive lobular carcinoma is distinguished by a unique vascular-immune microenvironment characterized by IL33-high differentiated endothelium, reduced angiogenic sprouting, and an enrichment of IL1RL1-expressing mast cells, which collectively differentiate it from invasive ductal carcinoma.
ViroNEXT is a fully automated, open-source multilayer hybrid pipeline that combines reference-based and machine-learning approaches to achieve high-precision detection of both known and highly divergent viruses in metagenomic data while significantly reducing false-positive classifications.
This paper presents a hybrid deep learning-based computer vision framework that combines Median Filtering, Histogram Equalization, and the Watershed Algorithm to achieve superior segmentation and feature extraction for brain organoid and embryoid body images.