Dimensionality Reduction Analysis Research Articles

Lightweight Noise Robust Spoofing Attack Detection using Cochleagram and ResNet Amalgamated Features

Abstract Speaker verification, like other biometric technologies, is vulnerable to spoofing attacks. An attacker impersonates a specific target speaker using impersonation, replay, Text-to-Speech (TTS), or Voice conversion (VC) techniques to gain unauthorized access to the system. The work in this paper, proposes a solution that uses an amalgamation of Cochleagram and Residual Network (ResNet) to implement the frontend feature extraction phase of an Audio Spoof Detection (ASD) system. Cochleagram generation, feature extraction-dimensionality reduction and classification are the three main phases of the proposed ASD system. In the first phase, the recorded audios have been converted into Cochleagrams by using Equivalent Rectangular Bandwidth (ERB) based gammatone filters. In the next phase, three variants of Residual Network (ResNet), ResNet50, ResNet41 and ResNet27, one by one, have been used for extracting dynamic features and Resnet50, ResNet41 that are fed to Linear Decrement Analysis (LDA) for dimensionality reduction. At last, in the classification phase, the LDA reduced features have been used for training four different machine learning classifiers such as Random Forest, Naïve Bayes, K-Nearest Neighbour (KNN), and eXtreme Gradient Boosting (XGBoost), individually. The proposed work in this paper concentrates on synthetic, replay, and deepfake attacks. The state-of-the-art ASVspoof 2019 Logical Access (LA), Physical Access (PA), Voice Spoofing Detection Corpus (VSDC) and DEepfake CROss-lingual (DECRO) datasets are utilised for training and testing the proposed ASD system. Additionally, we have assessed the performance of our proposed system under the influence of additive noise. Airplane noise at different SNR rate (0, dB 5dB, 10dB and -5dB) has been added to training and testing audios for the same. From the obtained results, it can be concluded that combination of Cochleagram and ResNet50 with XGBoost classifier outperforms all other implemented systems for detecting fake audios under noisy environment.

NipalsMCIA: Flexible Multi-Block Dimensionality Reduction in R via Nonlinear Iterative Partial Least Squares.

With the increased reliance on multi-omics data for bulk and single cell analyses, the availability of robust approaches to perform unsupervised learning for clustering, visualization, and feature selection is imperative. We introduce nipalsMCIA, an implementation of multiple co-inertia analysis (MCIA) for joint dimensionality reduction that solves the objective function using an extension to Non-linear Iterative Partial Least Squares (NIPALS). We applied nipalsMCIA to both bulk and single cell datasets and observed significant speed-up over other implementations for data with a large sample size and/or feature dimension. nipalsMCIA is available as a Bioconductor package at https://bioconductor.org/packages/release/bioc/html/nipalsMCIA.html, and includes detailed documentation and application vignettes. Supplementary Materials are available online.

MACHINE LEARNING APPLICATIONS IN RURAL HEALTHCARE: PREDICTIVE MODELING FOR IMMUNIZATION COMPLETION RATES IN OGUN STATE, NIGERIA

This study investigated the application of machine learning techniques for predicting child immunization completion in Ado-Odo/Ota Local Government Area, Ogun State, Nigeria, utilizing data from 8,808 immunization records across 15 primary healthcare centers. Using a quantitative research methodology with retrospective data analysis, we developed and compared predictive models for immunization completion patterns. Three machine learning algorithms were employed based on their proven effectiveness in healthcare applications: Logistic Regression for its interpretability in clinical settings, Support Vector Machine (SVM) for handling non-linear relationships in health data, and K-Nearest Neighbours (KNN) for processing demographic variables. The study analyzed immunization completion rates using these algorithms within a comprehensive framework incorporating Principal Component Analysis for dimensionality reduction. The Logistic Regression model demonstrated superior performance with 99.77% accuracy and an MSE of 0.0023, outperforming both SVM (99.32% accuracy) and KNN (99.03% accuracy) models. Notably, socioeconomic analysis revealed an unexpected pattern where high-income households showed lower immunization completion rates compared to low and moderate-income groups. The study's findings provide valuable insights for healthcare policy development and resource allocation strategies while demonstrating the practical applicability of machine learning in enhancing immunization program effectiveness in developing nations.

Investigation of T lymphocyte subsets in children with Mycoplasma pneumoniae pneumonia.

This study aims to characterize the majority of immune cell subsets in peripheral blood mononuclear cells in children with Mycoplasma pneumoniae pneumonia (MPP) by a 21-color flow cytometry panel. Patients who met the predetermined eligibility criteria for pneumonia diagnosis were recruited for the research study. Multi-color flow cytometry was conducted on the peripheral blood mononuclear cells of each patient group, which were then subjected to dimensionality reduction and cluster analysis. In our study, the proportion of activated CD4 + T cell and naïve CD8 + T in children with MPP was higher than that of children with non-MPP, and the proportion of CD8 + T cell and central memory CD8 + T cell in MPP children was lower. Central memory CD4 + T cell and activated CD4 + T cell in the severe MPP were higher than those in the mild MPP. The highest proportions of CD8 + T cell, CD8 + Tn cell, activated CD8 + T cell, and total activated T cell were observed in the pulmonary consolidation-mucous group when compared to the pulmonary consolidation-necrosis and bronchiolitis groups. In the pulmonary consolidation-necrosis group, the proportions of central memory CD4 + T cell and T helper 17 cell were higher than those in pulmonary consolidation-mucous and bronchiolitis groups. In the bronchiolitis group, the percentages of CD4 + T cell, naïve CD4 + T cell, and T helper 2 cell were higher than those in pulmonary consolidation-mucous and the pulmonary consolidation-necrosis groups. The T lymphocyte subsets were different among various groups, offering new insights into the immune system of pediatric patients with Mycoplasma pneumoniae pneumonia.

An Enhanced Gas Sensor Data Classification Method Using Principal Component Analysis and Synthetic Minority Over-Sampling Technique Algorithms.

This study addresses the challenge of multi-dimensional and small gas sensor data classification using a gelatin-carbon black (CB-GE) composite film sensor, achieving 91.7% accuracy in differentiating gas types (ethanol, acetone, and air). Key techniques include Principal Component Analysis (PCA) for dimensionality reduction, the Synthetic Minority Over-sampling Technique (SMOTE) for data augmentation, and the Support Vector Machine (SVM) and K-Nearest Neighbor (KNN) algorithms for classification. PCA improved KNN and SVM classification, boosting the Area Under the Curve (AUC) scores by 15.7% and 25.2%, respectively. SMOTE increased KNN's accuracy by 2.1%, preserving data structure better than polynomial fitting. The results demonstrate a scalable approach to enhancing classification accuracy under data constraints. This approach shows promise for expanding gas sensor applicability in fields where data limitations previously restricted reliability and effectiveness.

Automated analysis of ultrastructure through large-scale hyperspectral electron microscopy

Microscopy is a key technique to visualize and understand biology. Electron microscopy (EM) facilitates the investigation of cellular ultrastructure at biomolecular resolution. Cellular EM was recently revolutionized by automation and digitalisation allowing routine capture of large areas and volumes at nanoscale resolution. Analysis, however, is hampered by the greyscale nature of electron images and their large data volume, often requiring laborious manual annotation. Here we demonstrate unsupervised and automated extraction of biomolecular assemblies in conventionally processed tissues using large-scale hyperspectral energy-dispersive X-ray (EDX) imaging. First, we discriminated biological features in the context of tissue based on selected elemental maps. Next, we designed a data-driven workflow based on dimensionality reduction and spectral mixture analysis, allowing the visualization and isolation of subcellular features with minimal manual intervention. Broad implementations of the presented methodology will accelerate the understanding of biological ultrastructure.

Variants of KLF5 and KLF12 were related to endometrial cancer risk in the Chinese population

ABSTRACT Objectives Krüppel‑like factors (KLFs) are implicated in the progression of endometrial cancer (EC). This present study explored the correlation between variants of KLF5, KLF12 and EC risk in the Chinese population. Methods The Agena MassARRAY technology platform was utilized to genotype six single nucleotide polymorphisms (SNPs) in KLF5 and KLF12 genes among 509 women diagnosed with EC and 506 age-matched healthy women. Subsequently, the relationship between SNPs in KLF5 and KLF12 and EC risk was calculated using logistic regression analysis. The interactions between SNPs in KLF5 and KLF12 were analyzed to predict EC risk using multifactor dimensionality reduction (MDR) analysis. Results KLF12 rs12429889 was significantly associated with EC risk (codominant: OR = 1.53, p = 0.003; dominant: OR = 1.54, p = 0.004). In addition, rs7329599 was significantly associated with EC risk in participants aged ≤55 years (codominant: OR = 0.63, p = 0.014; dominant: OR = 0.67, p = 0.024), whereas rs12429889 was significantly associated with EC risk in participants aged >55 years (codominant: OR = 1.98, p = 0.004; dominant: OR = 2.06, p = 0.002). Conclusion Our findings revealed a significant correlation between KLF12 rs12429889 and rs7329599 and EC risk, highlighting their potential as diagnostic biomarkers.

Carbon meets credit: enhancing corporate ratings using principal component analysis and Gaussian mixture models

This study integrates environmental factors, such as carbon emissions and climate-related risks, with traditional financial metrics to enhance corporate credit rating predictions for approximately 2,000 publicly listed Chinese firms. Utilizing Principal Component Analysis (PCA) for dimensionality reduction and Gaussian Mixture Models (GMM) for probabilistic clustering, the approach effectively addresses multicollinearity and captures nuanced financial-environmental relationships. The PCA-GMM framework not only improves predictive accuracy over traditional methods but also maintains high interpretability, making it suitable for complex risk profiles. Empirical results demonstrate that firms with lower carbon emissions and robust financial health receive higher credit ratings, providing valuable insights for investors and policymakers. This model supports sustainable economic growth in alignment with China’s dual carbon goals by highlighting the critical interplay between financial performance and environmental responsibility.

Grasserie disease detection in silkworm using machine learning methods

Grasserie disease, caused by the Bombyx mori nuclear polyhedrosis virus (BmNPV), is one of the most severe diseases affecting silkworms, leading to significant economic losses in sericulture. This study leverages machine learning techniques to detect Grasserie disease in FC1 x FC2 double hybrid silkworms, using image processing and classification methods to automate detection. A dataset of 737 images, consisting of 322 infected and 415 healthy silkworms, was analyzed. Local Binary Patterns (LBP) were used for feature extraction, followed by Principal Component Analysis (PCA) for dimensionality reduction. A decision tree classifier achieved an accuracy of 92.1%, recall of 92.4%, precision of 90.2%, F1 score of 91.6%, and specificity of 90.7%. This paper outlines the methodology, discusses the results, and suggests future improvements for practical applications in sericulture.

Squirrel Search Optimization-based near-duplicate image detection

ABSTRACT Near duplicate (ND) image detection is a significant issue in a modern online environment with a wide range of applications like the detection of copyright violations and saving of storage space. Several existing ND detection techniques are perhaps not suitable for online applications due to the large computational burden, and may not successfully detect NDs containing large smooth and plain regions. In addition, the K-means algorithm used in most of the existing methods yield sub-optimal quantization of visual words. This article employs a robust algorithm of Squirrel Search Optimization (SSO) for quantization, fast-hessian matrix-based detector (FHMBD) and FAST Corner Detector (FCD) for the detection of KPs at both plain and non-smooth regions of all images, SURF for computing descriptors and Principal Component Analysis (PCA) for dimensionality reduction. The results of the developed method presented on five image databases. The proposed method offered 99.9% accuracy, 98.67% sensitivity, and 99.91% specificity respectively.

A composite Bayesian optimisation framework for material and structural design

In this contribution, a new design framework leveraging Bayesian optimisation is developed to enhance the efficiency and quality of material and structural design processes. The proposed framework comprises two main steps. The first step involves efficiently exploring the design space with a minimum number of sampled points to mitigate computational costs. In the subsequent step, a composite Bayesian optimisation strategy is employed to evaluate the objective function and identify the next candidate for sampling. By building a surrogate model for numerical simulation responses in a fixed-size latent response space and using techniques like Principal Component Analysis for dimensionality reduction, the framework effectively exploits the composition aspect of the objective function. Unlike traditional methods that rely on random sampling across the design space, our Bayesian optimisation approach uses a dynamic, adaptive sampling strategy. This method significantly reduces the number of required experiments while effectively managing uncertainty. We evaluate the framework’s performance across various design scenarios and conduct a critical comparative analysis against well-established data-driven approaches. These scenarios include linear and nonlinear material and structural behaviours, addressing multi-objective optimisation and data variability. Our findings demonstrate substantial improvements in performance and quality, particularly in nonlinear settings. This underscores the framework’s potential to advance design methodologies in material and structural engineering.

B-cell dynamics underlying poor response upon split-inactivated influenza virus vaccination.

This investigation elucidated the differences in humoral and H1N1 HA-specific memory B-cells response in participants exhibiting distinct immune response patterns prior to and after vaccination with Fluzone, the quadrivalent split-inactivated seasonal influenza virus vaccine. Participants were categorized into persistent non-responders and persistent responders based on their hemagglutination-inhibition (HAI) antibody titers to the H1N1 component from each vaccine administered between the 2019-2020 to 2023-2024 seasons. Persistent responders had higher fold change in H1N1 HA-specific CD21 expressing B-cells, plasmablasts, and plasma cells. A significant increase in H1N1 HA-specific transitional B-cells in persistent non-responders was observed. The frequency and fold change of H1N1-specific IgM-expressing memory B-cells was higher in persistent non-responders. Dimensionality reduction analysis also demonstrated higher IgM expression for persistent non-responders than persistent responders. Furthermore, persistent non-responders had a significant fold change increase in IgA tissue-like memory, IgG exhausted tissue-like memory, and double negative (DN) activated memory cells. In contrast, persistent responders had increased frequency of IgG-activated memory B-cells, IgG resting B-cells and DN resting B-cells. Correlation analysis revealed a positive correlation between HAI titers and DN memory B-cells and a negative correlation between HAI titers and IgG-expressing memory B-cells in persistent non-responders. Conversely, persistent responders had a positive correlation between HAI titers and IgA resting memory B-cells and a negative correlation between IgG memory B-cells and DN memory B-cells. Overall, this study provided valuable insights into the differential immune memory B-cell responses following influenza virus vaccination and paves the way for future research to further unravel the complexities of vaccine-induced memory B-cells and ultimately improve vaccination strategies against influenza virus infection.

A Method of Intelligent Driving-Style Recognition Using Natural Driving Data

At present, achieving efficient, sustainable, and safe transportation has led to increasing attention on driving behavior recognition and advancements in autonomous driving. Identifying diverse driving styles and corresponding types is crucial for providing targeted training and assistance to drivers, enhancing safety awareness, optimizing driving costs, and improving autonomous driving systems responses. However, current studies mainly focus on specific driving scenarios, such as free driving, car-following, and lane-changing, lacking a comprehensive and systematic framework to identify the diverse driving styles. This study proposes a novel, data-driven approach to driving-style recognition utilizing naturalistic driving data NGSIM. Specifically, the NGSIM dataset is employed to categorize car-following and lane-changing groups according to driving-state extraction conditions. Then, characteristic parameters that fully represent driving styles are optimized through correlation analysis and principal component analysis for dimensionality reduction. The K-means clustering algorithm is applied to categorize the car-following and lane-changing groups into three driving styles: conservative, moderate, and radical. Based on the clustering results, a comprehensive evaluation of the driving styles is conducted. Finally, a comparative evaluation of SVM, Random Forest, and KNN recognition indicates the superiority of the SVM algorithm and highlights the effectiveness of dimensionality reduction in optimizing characteristic parameters. The proposed method achieves over 97% accuracy in identifying car-following and lane-changing behaviors, confirming that the approach based on naturalistic driving data can effectively and intelligently recognize driving styles.

A Biomechanical Dataset of 1,798 Healthy and Injured Subjects During Treadmill Walking and Running

Quantitative biomechanical gait analysis is an important clinical and research tool for injury and disease diagnosis and treatment. However, one major criticism is that gait analysis laboratories largely operate in isolation and there is a lack of benchmark datasets, which can be used to advance research and statistical methodologies. To address this, we present an open biomechanics dataset of n = 1798 healthy and injured, young and older adults during treadmill walking and/or running at a range of gait speeds. The full dataset is available on Figshare+ and data files are contained within a series of zipped folders with folder names representing the subject ID. Each subject ID folder contains walking and/or running data containing raw marker trajectory data along with metadata for each participant. Five tutorials are also provided, demonstrating aspects such as loading data files, sample analyses of discrete variables, and calculating joint angles from code along with covering more complex topics such as principal component analysis for dimensionality reduction, statistical parametric mapping, and conducting unsupervised clustering.

BIOM-67. DIFFERENTIAL DNA METHYLATION PATTERNS IN THE CSF OF PATIENTS WITH DIFFUSE GLIOMAS

Abstract Studies have shown that different central nervous system (CNS) tumor types exhibit unique DNA methylation profiles. DNA methylation-based machine learning classification of CNS tumor tissue has been demonstrated to be highly-accurate in differentiating between tumor types, including between diffuse glioma subtypes. While similar analyses have been performed using cell-free DNA (cfDNA) from blood, few studies have looked at the potential of DNA methylation as a biomarker for diffuse gliomas using CSF-derived cfDNA. In this study, we analyzed CSF cfDNA methylation patterns in diffuse glioma patients. DNA methylation profiling of CSF of patients with glioblastoma, IDH-wildtype (GBM) (n=37), IDH-mutant diffuse gliomas (n=12), and non-neoplastic conditions (n=5) was performed using cell-free methylated DNA immunoprecipitation and high-throughput sequencing (cfMeDIP-seq). Methylation quantification was estimated for each genomic region. Significance testing was performed between diagnostic groups to discover differentially-methylated regions (DMRs). The 300 most differentially-methylated regions (DMRs) for each comparison (IDH-wildtype vs. IDH-mutant, IDH-mutant vs. control, IDH-wildtype vs. control), for a total of 900 DMRs, were utilized as features in dimensionality reduction analysis and machine learning. Out of ~1×107 defined genomic regions, there were ~7×105 DMRs between GBM and control samples and ~1.5×106 DMRs between GBM and IDH-mutant samples, accounting for ~7% and ~15% of the genome, respectively. Uniform manifold approximation and projection (UMAP) and hierarchical clustering analysis revealed that samples were mostly clustered by diagnosis. Random forest analysis of the 900 DMRs with an 80/20 train/test split resulted in an overall accuracy of 0.815. In conclusion, CSF cfDNA methylation profiles of diffuse glioma patients exhibit differential methylation patterns between IDH-mutant and IDH-wildtype tumors, as well as between tumor and control, and could be informative biomarkers for the diagnosis of patients with diffuse gliomas.

Wearable device data-driven athlete injury detection and rehabilitation monitoring algorithm

In recent years, sports wearable technology has completely changed the way athletes prepare, compete, and recover. Wearable technology has a lot to offer in the rehabilitation process, which is essential to an athlete’s return to their best performance. Wearable devices for athlete injury detection pose potential challenges like data quality, security, and privacy, impacting accuracy, reliability, and effectiveness. To solve these problems, an innovative injury detection and rehabilitation monitoring (IDRM) system was proposed for athletes. By employing an adjustable recurrent neural network (ARNN) to detect anomalies in injury risks such as abnormal joint movements in athletes. In this study, biomechanics data was collected from sports athletes through wearable devices, and the wearable system provided feedback to the user. A redefined convolutional neural network (RCNN) was utilized to monitor the rehabilitation process. This system tracks athlete’s rehabilitation progress and ensures that progress monitors were performed correctly, and the system, feasibility was evaluated on 10 healthy subjects performing 4 different rehabilitation exercises. Each exercise was performed four times monitoring and validation. The data was preprocessed using a Gaussian filter to remove noise from the obtained data. Then the features are extracted using independent component analysis (ICA) for dimensionality reduction from preprocessed data. The proposed method is implemented using Python software. In comparative analysis, the performance of ARNN showed high performance, with an F1-measure of 91.6%, accuracy of 93.5%, recall of 92.8%, and precision of 91.4%. With a 95% accuracy rate, 98% F1 measure, 94% precision, and 93% recall, the RCNN model functioned effectively. The result showed the proposed method achieved better performance in athlete injury detection and accurately recognizing all rehabilitation monitoring. This study provides a complete approach to athlete health management by highlighting the integration of rehabilitation monitoring and injury detection into an overall structure.

Analysis and knowledge extraction of newborn resuscitation activities from annotation files.

Deprivation of oxygen in an infant during and after birth leads to birth asphyxia, which is considered one of the leading causes of death in the neonatal period. Adequate resuscitation activities are performed immediately after birth to save the majority of newborns. The primary resuscitation activities include ventilation, stimulation, drying, suction, and chest compression. While resuscitation guidelines exist, little research has been conducted on measured resuscitation episodes. Objective data collected for measuring and registration of the executed resuscitation activities can be used to generate temporal timelines. This paper is primarily aimed to introduce methods for analyzing newborn resuscitation activity timelines, through visualization, aggregation, redundancy and dimensionality reduction. We are using two datasets: 1) from Stavanger University Hospital with 108 resuscitation episodes, and 2) from Haydom Lutheran Hospital with 76 episodes. The resuscitation activity timelines were manually annotated, but in future work we will use the proposed method on automatically generated timelines from video and sensor data. We propose an encoding generator with unique codes for combination of activities. A visualization of aggregated episodes is proposed using sparse nearest neighbor graph, shown to be useful to compare datasets and give insights. Finally, we propose a method consisting of an autoencoder trained for reducing redundancy in encoded resuscitation timeline descriptions, followed by a neighborhood component analysis for dimensionality reduction. Visualization of the resulting features shows very good class separability and potential for clustering the resuscitation files according to the outcome of the newborns as dead, admitted to NICU or normal. This shows great potential for extracting important resuscitation patterns when tested on larger datasets.

Optimum Combination of Spectral Variables for Crop Mapping in Heterogeneous Landscapes based on Sentinel-2 Time Series and Machine Learning

Abstract. This article aimed to determine a workflow for more efficient large-scale crop mapping using a time series of images from the Sentinel-2 Satellite, statistical methods of attribute selection, and machine learning. The proposed methodology explores the best possible combination of spectral variables related to vegetation (16 vegetation indices in the RGB, NIR, SWIR, and Red Edge regions) to characterize different spectro-temporal profiles of Land Use and Land Cover (LULC) in spatially heterogeneous landscapes. First, we applied a data dimensionality reduction analysis using the PCA (Principal Component Analysis) method. Subsequently, the variables that showed the highest statistical correlation between each other were used in the spectro-temporal classification process, using the Random Forest, TempCNN, and LightTAE algorithms, following three different strategies: C1 (ALL), C2 (BE + IV (Red Edge)) and C3 (BE + IV (without Red Edge)), where ALL – All variables; BE – Spectral Bands; IV – Vegetation Indices. Given the results found, the C2 classification scenario (with bands B3, B4, B5, B6, B7, B8, and B8A and the NDRE1, RESI, and MSR indexes) demonstrated the best LULC classification accuracy at the crop pattern level, compared to the other scenarios, with average values of 0.91, 0.88, 0.91, 0.89, and 0.89 (Global Accuracy, Producer Accuracy, User Accuracy, Kappa index, and F1-Score, respectively, for the TempCNN model), the which emphasized the importance of both qualitative and quantitative variability of sampling data and variables based on the Red Edge region for improving LULC classification processes in large-scale heterogeneous landscapes.

Spatial Transcriptomics of IPMN Reveals Divergent Indolent and Malignant Lineages.

Intraductal papillary mucinous neoplasms (IPMN) occur in 5-10% of the population, but only a small minority progress to pancreatic ductal adenocarcinoma (PDAC). The lack of accurate predictors of high-risk disease leads both to unnecessary operations for indolent neoplasms as well as missed diagnoses of PDAC. Digital spatial RNA profiling (DSP-RNA) provides an opportunity to define and associate transcriptomic states with cancer risk. Whole-transcriptome DSP-RNA profiling was performed on 10 IPMN specimens encompassing the spectrum of dysplastic changes from normal duct to cancer. Ductal epithelial regions within each tissue were annotated as normal duct (NL), low-grade dysplasia (LGD), high-grade dysplasia (HGD), or invasive carcinoma (INV). Gene expression count data was generated by Illumina sequencing and analyzed with R/Bioconductor. Dimension reduction analysis exposed three clusters reflecting IPMN transcriptomic states denoted "normal-like" ( cNL ), "low-risk" ( cLR ) and "high-risk" ( cHR ). In addition to specific marker genes, the three states exhibited significant enrichment for the exocrine, classical, and basal-like programs in PDAC. Specifically, exocrine function diminished in cHR , classical activation distinguished neoplasia from cNL , and basal-like genes were specifically upregulated in cHR . Intriguingly, markers of cHR were detected in NL and LGD regions from specimens with PDAC but not low-grade IPMN. DSP-RNA of IPMN revealed low-risk (indolent) and high-risk (malignant) expression programs that correlated with the activity of exocrine and basal-like PDAC signatures, respectively, and distinguished pathologically low-grade from malignant specimens. These findings contextualize IPMN pathogenesis and have the potential to transform existing risk stratification models. Current consensus guidelines for management of intraductal papillary mucinous neoplasms (IPMN) of the pancreas utilize clinical and radiographic criteria for risk stratification. Unfortunately, the estimated positive predictive value of these criteria for IPMN-associated pancreatic ductal adenocarcinoma (PDAC) is under 50%, indicating that over half of pancreatectomies are performed for benign disease. Moreover, nearly 15% of patients who were deemed "low risk" by the same criteria harbored PDAC. Surgical resection of IPMN has maximal benefit when performed prior to the development of PDAC, as evidence of carcinoma has been associated with a high rate of recurrence and poor overall survival. Thus, the development of molecular diagnostics that improve the accuracy of IPMN risk classification would have immediate relevance for patient care, both in terms of better selecting patients for potentially curative operations, as well as sparing patients with low-risk lesions from invasive procedures.

SCRN: Single-cell Gene Regulatory Network Identification in Alzheimer's Disease.

Alzheimer's disease (AD) is the most common neurodegenerative disease, and it consumes considerable medical resources with increasing number of patients every year. Mounting evidence show that the regulatory disruptions altering the intrinsic activity of genes in brain cells contribute to AD pathogenesis. To gain insights into the underlying gene regulation in AD, we proposed a graph learning method, Single-Cell based Regulatory Network (SCRN), to identify the regulatory mechanisms based on single-cell data. SCRN implements the γ-decaying heuristic link prediction based on graph neural networks and can identify reliable gene regulatory networks using locally closed subgraphs. In this work, we first performed UMAP dimension reduction analysis on single-cell RNA sequencing (scRNA-seq) data of AD and normal samples. Then we used SCRN to construct the gene regulatory network based on three well-recognized AD genes (APOE, CX3CR1, and P2RY12). Enrichment analysis of the regulatory network revealed significant pathways including NGF signaling, ERBB2 signaling, and hemostasis. These findings demonstrate the feasibility of using SCRN to uncover potential biomarkers and therapeutic targets related to AD.