Two-dimensional nanomaterials-based optical biosensors empowered by machine learning for intelligent diagnosis

TrAC Trends in Analytical Chemistry, Год журнала: 2025, Номер unknown, С. 118162 - 118162

Опубликована: Фев. 1, 2025

Язык: Английский

---

Unbiasedly decoding the tumor microenvironment with single-cell multiomics analysis in pancreatic cancer

Molecular Cancer, Год журнала: 2024, Номер 23(1)

Опубликована: Июль 9, 2024

Abstract Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy with poor prognosis and limited therapeutic options. Research on the tumor microenvironment (TME) of PDAC has propelled development immunotherapeutic targeted strategies promising future. The emergence single-cell sequencing mass spectrometry technologies, coupled spatial omics, collectively revealed heterogeneity TME from multiomics perspective, outlined trajectories cell lineages, important functions previously underrated myeloid cells stroma cells. Concurrently, these findings necessitated more refined annotations biological at cluster or level. Precise identification all clusters urgently needed to determine whether they have been investigated adequately identify target antitumor potential, design compatible treatment strategies, resistance. Here, we summarize recent research level, an unbiased focus potential classification bases every cellular component within TME, look forward prospects integrating data retrospectively reusing bulk data, hoping provide new insights into TME.

Язык: Английский

---

From High Dimensions to Human Insight: Exploring Dimensionality Reduction for Chemical Space Visualization

Molecular Informatics, Год журнала: 2024, Номер 44(1)

Опубликована: Дек. 5, 2024

Abstract Dimensionality reduction is an important exploratory data analysis method that allows high‐dimensional to be represented in a human‐interpretable lower‐dimensional space. It extensively applied the of chemical libraries, where structure ‐ as feature vectors‐are transformed into 2D or 3D space maps. In this paper, commonly used dimensionality techniques Principal Component Analysis (PCA), t‐Distributed Stochastic Neighbor Embedding (t‐SNE), Uniform Manifold Approximation and Projection (UMAP), Generative Topographic Mapping (GTM) are evaluated terms neighborhood preservation visualization capability sets small molecules from ChEMBL database.

Язык: Английский

---

Diffusive topology preserving manifold distances for single-cell data analysis

Proceedings of the National Academy of Sciences, Год журнала: 2025, Номер 122(4)

Опубликована: Янв. 24, 2025

Manifold learning techniques have emerged as crucial tools for uncovering latent patterns in high-dimensional single-cell data. However, most existing dimensionality reduction methods primarily rely on 2D visualization, which can distort true data relationships and fail to extract reliable biological information. Here, we present DTNE (diffusive topology neighbor embedding), a framework that faithfully approximates manifold distance enhance cellular dynamics. constructs matrix using modified personalized PageRank algorithm, thereby preserving topological structure while enabling diverse analyses. This approach facilitates distribution-based relationship analysis, pseudotime inference, clustering within unified framework. Extensive benchmarking against mainstream algorithms datasets demonstrates DTNE’s superior performance maintaining geodesic distances revealing significant patterns. Our results establish powerful tool analysis meaningful insights.

Язык: Английский

---

Investigation of cell development and tissue structure network based on natural Language processing of scRNA-seq data

Journal of Translational Medicine, Год журнала: 2025, Номер 23(1)

Опубликована: Март 4, 2025

Single-cell multi-omics technologies, particularly single-cell RNA sequencing (scRNA-seq), have revolutionized our understanding of cellular heterogeneity and development by providing insights into gene expression at the level. Investigating influence genes on behavior is crucial for elucidating cell fate determination differentiation, processes, disease mechanisms. Inspired NLP, we present a novel scRNA-seq analysis method that treats as analogous to words. Using word2vec embed sequences derived from networks, generate vector representations genes, which are then used represent cells summing vectors subsequently tissues aggregating vectors. Our NLP-based approach analyzes data generating cells, tissues. This multi-scale includes mapping states in space reveal developmental trajectories, quantifying similarity using Euclidean distance, constructing inter-tissue relationship networks aggregated offers computationally efficient analyzing embedding similar those large language model pre-training, but without requiring high-performance computing clusters. By embeddings capture functional relationships, this facilitates study impact perturbations, clustering, construction tissue networks. provides valuable tool analysis.

Язык: Английский

---

LocPro: a deep learning-based prediction of protein subcellular localization for promoting multi-directional pharmaceutical research

Journal of Pharmaceutical Analysis, Год журнала: 2025, Номер unknown, С. 101255 - 101255

Опубликована: Март 1, 2025

Язык: Английский

---

Artificial Intelligence-Powered Surface-Enhanced Raman Spectroscopy for Biomedical Applications

Analytical Chemistry, Год журнала: 2025, Номер unknown

Опубликована: Март 27, 2025

Язык: Английский

---

Lipidome visualisation, comparison, and analysis in a vector space

PLoS Computational Biology, Год журнала: 2025, Номер 21(4), С. e1012892 - e1012892

Опубликована: Апрель 15, 2025

A shallow neural network was used to embed lipid structures in a 2- or 3-dimensional space with the goal that structurally similar species have vectors. Tests on complete databanks show method automatically produces distributions which follow conventional classifications. The embedding is accompanied by web-based software, Lipidome Projector. This displays user lipidomes as 2D 3D scatterplots for quick exploratory analysis, quantitative comparison and interpretation at structural level. Examples of published data sets were qualitative literature interpretation.

Язык: Английский

---

Interpretability study of earthquake-induced landslide susceptibility combining dimensionality reduction and clustering

Frontiers in Earth Science, Год журнала: 2025, Номер 13

Опубликована: Апрель 25, 2025

An earthquake of magnitude Ms5.8 struck Barkam City, Aba Prefecture, Sichuan Province, China, on the morning 10 June 2022. This was followed by two additional earthquakes magnitudes Ms6.0 and Ms5.2. The triggered significant geological hazards, impacting City surrounding areas. Using Random Forest (RF) Extreme Gradient Boosting (XGBoost) machine learning models, we assessed landslide susceptibility in identified key influencing factors. study applied SHAP method to evaluate importance various factors, used UMAP for dimensionality reduction, employed HDBSCAN clustering algorithm classify data, thereby enhancing interpretability models. results show that XGBoost outperforms RF terms accuracy, precision, recall, F1 score, KC, MCC. primary factors occurrence are topographic features, seismic activity, precipitation intensity. research not only introduces innovative techniques methods analysis but also provides a scientific foundation emergency response post-disaster planning related risks following City.

Язык: Английский

---

From High Dimensions to Human Comprehension: Exploring Dimensionality Reduction for Chemical Space Visualization

Опубликована: Авг. 16, 2024

Dimensionality reduction is an important exploratory data analysis method that allows high-dimensional to be represented in a human-interpretable lower-dimensional space. It extensively applied the of chemical libraries, where structure — as feature vectors—are transformed into 2D or 3D space maps. In this paper, commonly used dimensionality techniques Principal Component Analysis (PCA), t-Distributed Stochastic Neighbor Embedding (t-SNE), Uniform Manifold Approximation and Projection (UMAP), Generative Topographic Mapping (GTM) are evaluated for exploration subsets small molecule organic compounds from ChEMBL database. The performance these methods examined terms neighborhood preservation visualization capabilities, strengths limitations discussed.

Язык: Английский

---