Large-scale Data Analysis Research Articles

Content Modeling in Multi-Platform Multilingual Social Media Data

An increase in the use of social media as the primary news source for the general population has created an ecosystem in which organic conversation commingles with inorganically seeded and amplified narratives, which can include public relations and marketing activity but also covert and malign influence operations. An efficient and easily understandable analysis of such data is important, as it allows relevant stakeholders to protect online communities and free discussion while better identifying activity and content that may violate social media platform terms of service. To accomplish this, we propose a method of large-scale social media data analysis, which allows for multilingual conversations to be analyzed in depth across any number of social media platforms simultaneously. Our method uses a text embedding model, i.e., a natural language processing model that holds semantic and contextual understandings of language. The model uses an “understanding” of language to represent posts as coordinates in a high-dimensional space, such that posts with similar meanings are assigned coordinates close together. We then cluster and analyze the posts to identify online topics of conversation existing across multiple social media platforms. We explicitly show how our method can be applied to four different datasets, three consisting of Chinese social media posts related to the Belt and Road Initiative and one relating to the Russia-Ukraine war, and we find politically-influenced conversations that contain misleading information relating to the Chinese government and the Russia-Ukraine war.

RMATS-turbo: an efficient and flexible computational tool for alternative splicing analysis of large-scale RNA-seq data.

Pre-mRNA alternative splicing is a prevalent mechanism for diversifying eukaryotic transcriptomes and proteomes. Regulated alternative splicing plays a role in many biological processes, and dysregulated alternative splicing is a feature of many human diseases. Short-read RNA sequencing (RNA-seq) is now the standard approach for transcriptome-wide analysis of alternative splicing. Since 2011, our laboratory has developed and maintained Replicate Multivariate Analysis of Transcript Splicing (rMATS), a computational tool for discovering and quantifying alternative splicing events from RNA-seq data. Here we provide a protocol for the contemporary version of rMATS, rMATS-turbo, a fast and scalable re-implementation that maintains the statistical framework and user interface of the original rMATS software, while incorporating a revamped computational workflow with a substantial improvement in speed and data storage efficiency. The rMATS-turbo software scales up to massive RNA-seq datasets with tens of thousands of samples. To illustrate the utility of rMATS-turbo, we describe two representative application scenarios. First, we describe a broadly applicable two-group comparison to identify differential alternative splicing events between two sample groups, including both annotated and novel alternative splicing events. Second, we describe a quantitative analysis of alternative splicing in a large-scale RNA-seq dataset (~1,000 samples), including the discovery of alternative splicing events associated with distinct cell states. We detail the workflow and features of rMATS-turbo that enable efficient parallel processing and analysis of large-scale RNA-seq datasets on a compute cluster. We anticipate that this protocol will help the broad user base of rMATS-turbo make the best use of this software for studying alternative splicing in diverse biological systems.

Two different autorefractors for vision screening in children and adolescents.

To compare the consistency of two autorefractors (Tianle RM-9000 and Topcon KR-800) for school-age myopia children, and to provide a basis for largescale data analysis and comparison. The refractive error in 909 subjects (age 4-18y) were measured using both autorefractors without cycloplegia. The data were analyzed using Fourier decomposition and the correlation coefficients, intraclass correlation coefficients (ICC), and Bland-Altman limits of agreement (LoA) for each parameter were calculated. There was a strong correlation between the spherical equivalent (SE), sphere diopter (DS), and cylinder diopter (DC) readings of the Tianle RM-9000 and those of the Topcon KR-800, with correlation coefficient values of 0.98, 0.98 and 0.83 and ICC values of 0.99, 0.99 and 0.93, respectively. However, the correlation coefficients and ICC values of J0 and J45 were unreliable (R=-0.004, -0.034; both ICC<0.10). Bland-Altman analysis revealed that SE, DS, and DC measured by the Tianle RM-9000 were significantly biased toward myopia compared with the Topcon KR-800, and the mean differences were -0.072, -0.026, -0.091 D, respectively (all P<0.01). The minimum absolute value of the difference within the 95% LoA for SE, DS, and DC was 0.63 D, 0.50 D, 0.62 D, respectively; all these values were in the clinically acceptable range. For J0 and J45, the mean differences were close to zero (P=0.43, 0.84); however, the 95% LoA were relatively wide (J0 SD: 0.53; 95%CI: -1.00, 1.10; J45 SD: 0.52; 95%CI: -1.00, 1.00). The two autorefractors are consistent with each other, as the differences in SE, DS, and DC were within the clinically acceptable range. Readers can compare the data measured by either device in different studies and use the two devices in the same study to generate a dataset that can be analyzed together. However, the J0 and J45 vectors are unreliable and should not be used to assess astigmatism.

Clinical trials are becoming more complex: a machine learning analysis of data from over 16,000 trials

The past decade has seen substantial innovation in clinical trials, including new trial formats, endpoints, and others. Also there have been regulatory changes, increasing competitive pressures and other external factors which impact clinical trials. In parallel, trial timelines have increased and success rates remain stubbornly low. This has led many observers to question whether clinical trials have become overly complex and if this complexity is always needed. Here we present a large-scale analysis of protocols and other data from over 16,000 trials. Using a machine learning algorithm, we automatically assessed key features of these trials, such as number of endpoints, number of inclusion–exclusion criteria and others. Using a regression analysis we combined these features into a new metric, the Trial Complexity Score, which correlates with overall clinical trial duration. Looking at this score across different clinical phases and therapeutic areas we see substantial increases over time, suggesting that clinical trials are indeed becoming more complex. We discuss drivers of increasing trial complexity, necessary or helpful (‘good’) complexity versus unnecessary (‘bad’) complexity, and we explore mechanisms of how sponsors of clinical trials can reduce trial complexity where appropriate.

Retraction Note: Fuzzy clustering-based skyline query preprocessing algorithm for large-scale flow data analysis

Retraction Note: Fuzzy clustering-based skyline query preprocessing algorithm for large-scale flow data analysis

Boosting HPC data analysis performance with the ParSoDA-Py library

Developing and executing large-scale data analysis applications in parallel and distributed environments can be a complex and time-consuming task. Developers often find themselves diverted from their application logic to handle technical details about the underlying runtime and related issues. To simplify this process, ParSoDA, a Java library, has been proposed to facilitate the development of parallel data mining applications executed on HPC systems. It simplifies the process by providing built-in scalability mechanisms relying on the Hadoop and Spark frameworks. This paper presents ParSoDA-Py, the Python version of the ParSoDA library, which allows for further support of commonly used runtimes and libraries for big data analysis. After a complete library redesign, ParSoDA can be now easily integrated with other Python-based distributed runtimes for HPC systems, such as COMPSs and Apache Spark, and with the large ecosystem of Python-based data processing libraries. The paper discusses the adaptation process, which takes into consideration the new technical requirements, and evaluates both usability and scalability through some case study applications.

Dynamic state estimation of distribution networks with mixed measurements based on Cubature Kalman Filter

The integration of synchronized phasor measurement units (μPMU) has introduced diverse data sources into the distribution network. Each has distinct time scales and precision levels. To enable effective decision-making, it is crucial to amalgamate these measurements efficiently, ensuring both speed and accuracy in distribution network state estimation. This study explores a dynamic state estimation approach for distribution networks, employing Cubature Kalman Filtering (CKF) as the central methodology. By deeply fusing μPMU, AMI and SCADA data and incorporating CKF techniques, it improves accuracy and stability, which is validated on the IEEE 37-node system. This research supports advanced distribution network apps and large-scale data analysis.

LabelSeg: segment annotation for tumor copy number alteration profiles.

Somatic copy number alterations (SCNAs) are a predominant type of oncogenomic alterations that affect a large proportion of the genome in the majority of cancer samples. Current technologies allow high-throughput measurement of such copy number aberrations, generating results consisting of frequently large sets of SCNA segments. However, the automated annotation and integration of such data are particularly challenging because the measured signals reflect biased, relative copy number ratios. In this study, we introduce labelSeg, an algorithm designed for rapid and accurate annotation of CNA segments, with the aim of enhancing the interpretation of tumor SCNA profiles. Leveraging density-based clustering and exploiting the length-amplitude relationships of SCNA, our algorithm proficiently identifies distinct relative copy number states from individual segment profiles. Its compatibility with most CNA measurement platforms makes it suitable for large-scale integrative data analysis. We confirmed its performance on both simulated and sample-derived data from The Cancer Genome Atlas reference dataset, and we demonstrated its utility in integrating heterogeneous segment profiles from different data sources and measurement platforms. Our comparative and integrative analysis revealed common SCNA patterns in cancer and protein-coding genes with a strong correlation between SCNA and messenger RNA expression, promoting the investigation into the role of SCNA in cancer development.

Analysis of the Change of Accounting Firms by Listed Companies- -Take New Asia Manufacturing Process (Zhejiang) Co., Ltd. as an Example

In the current market economic environment, the phenomenon of frequent replacement of accounting firms in China has attracted wide attention inside and outside the industry. On the one hand, this trend reflects the change of market competition and corporate business needs. On the other hand, it may also affect investors 'decisions and the public's trust in listed companies. Therefore, it is obviously necessary to study the motivation behind the replacement of accounting firms in listed companies. This paper aims to explore the motivation of Chinese listed companies to change accounting firms and their potential influence. Firstly, through large-scale data collection and analysis, this paper described the basic status of changing accounting firms by listed companies in China in recent years.Then, we discussed the various possible reasons for listed companies to replace their accounting firms. After in-depth research and analysis, we find that although the replacement of accounting firms can enhance the credibility of the financial reports. Therefore, this paper put forward a series of regulatory countermeasures. The main significance of this study lies in the in-depth discussion of the phenomenon of listed companies changing accounting firms, which provides more detailed market data and rich empirical analysis framework for future research.

Abstract B114: Mutational "lineage traces" of an acinar cell-of-origin in human pancreatic ductal adenocarcinoma

Abstract A clinical challenge of pancreatic ductal adenocarcinoma (PDAC) is that most cases are detected late, after intractable metastasis is well underway. This also poses a challenge to our basic understanding of PDAC development, as the early steps of neoplasia are separated by a gulf of years or decades from accessible biospecimens and cell lines. Although mouse genetic models of PDAC have helped to fill in this gap, key questions remain incompletely answered. For example, what is the cell of origin of human PDAC – and does the answer to this question actually matter at a clinical level? In mouse, PDAC can be induced from both acinar and duct cells, with the former apparently more susceptible to transformation while the latter appears harder to transform, but capable of generating more aggressive tumors. Meanwhile, human PDAC has been divided into two major subtypes based on genomic data, classical and basal-like, the latter of which appears to be more aggressive. Is there a relationship between cell-of-origin, tumor subtype and clinical outcomes? This question requires the ability to directly infer cell-of-origin in human tumors, an approach to which we report here. Studies in human non-pancreatic cancers have identified somatic “passenger” (non-selected) mutations in several highly-expressed genes of the normal organ – for example, mutations in the ALB (albumin) gene in hepatocellular carcinoma, or SFTP genes (surfactant proteins) in lung adenocarcinoma. These might indicate a mutational process occurring in normal cells, prior to transformation, and constituting a “lineage mark” for the cell of origin – however, as these genes are also highly-transcribed in the cancer cells themselves, their mutations may arise after tumorigenesis is underway. This problem does not apply to acinar genes and PDAC: we and others have established that an initiating event in Kras-induced transformation of acinar cells is complete downregulation of acinar-specific gene products. Therefore, if a mechanism of transcription-induced mutation applies in the pancreas, we would expect it to leave a trace in acinar-derived PDAC. We have completed a large-scale analysis (n&amp;gt;2,000) of human whole genome sequencing data, across several carcinoma types including pancreas and liver, focused on unbiased detection of tissue-specific mutations similar to those previously found in the ALB and SFTP genes. Remarkably, we find robust enrichment of mutations in acinar-specific genes in PDAC, affecting approximately 25% of tumors. This constitutes the first in vivo genetic evidence for an acinar cell-of-origin in human pancreatic cancer, and opens the door to comparative genomic, transcriptomic and outcomes analyses of acinar-derived tumors. These are underway, and the results will be described in our presentation. Citation Format: Charles Murtaugh. Mutational "lineage traces" of an acinar cell-of-origin in human pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr B114.

Big data clustering using fuzzy based energy efficient clustering and MobileNet V2

Big data analytics (BDA) is a systematic way to analyze and detect various patterns, relationships, and trends in vast amounts of data. Big data analysis and processing require significant effort, techniques, and equipment. The Hadoop framework software uses the MapReduce approach to do large-scale data analysis using parallel processing in order to generate results as soon as possible. Due to the traditional algorithm’s longer execution time and difficulty in processing big amounts of data, this is one of the main issues. Clusters are highly correlated inside each other but are not highly correlated with one another. The technique of effectively allocating limited resources is known as an optimization algorithm for clustering. For processing large amounts of data with several dimensions, the conventional optimization approach is insufficient. By using a fuzzy method, this can be prevented. In this paper, we proposed Fuzzy based energy efficient clustering approach to enhance the clustering mechanism. In summary, Fuzzy based energy efficient clustering introduces a function that measures the distance between the cluster center and the instance, which aids in improved clustering, and we then present the MobileNet V2 model to improve efficiency and speed up computation. To enhance the method’s performance and reduce its time complexity, the distributed database simulates the shared memory space and parallelizes on the MapReduce framework on the Hadoop cloud computing platform. The proposed approach is evaluated using performance metrics such as Accuracy, Precision, Adjusted Rand Index (ARI), Recall, F1-Score, and Normalized Mutual Information (NMI). The experimental findings indicate that the proposed approach outperforms the existing techniques in terms of clustering accuracy.

V-pipe 3.0: a sustainable pipeline for within-sample viral genetic diversity estimation.

The large amount and diversity of viral genomic datasets generated by next-generation sequencing technologies poses a set of challenges for computational data analysis workflows, including rigorous quality control, scaling to large sample sizes, and tailored steps for specific applications. Here, we present V-pipe 3.0, a computational pipeline designed for analyzing next-generation sequencing data of short viral genomes. It is developed to enable reproducible, scalable, adaptable, and transparent inference of genetic diversity of viral samples. By presenting 2 large-scale data analysis projects, we demonstrate the effectiveness of V-pipe 3.0 in supporting sustainable viral genomic data science.

Integrated bioinformatics analysis of nucleotide metabolism based molecular subtyping and biomarkers in lung adenocarcinoma.

Lung adenocarcinoma (LUAD), a predominant subtype of non-small cell lung cancers, continues to challenge treatment outcomes due to its heterogeneity and complex tumor microenvironment (TME). Dysregulation in nucleotide metabolism has been identified as a significant factor in tumorigenesis, suggesting its potential as a therapeutic target. This study analyzed LUAD samples from The Cancer Genome Atlas (TCGA) using Non-negative Matrix Factorization (NMF) clustering, Weighted Correlation Network Analysis (WGCNA), and various machine learning techniques. We investigated the role of nucleotide metabolism in relation to clinical features and immune microenvironment through large-scale data analysis and single-cell sequencing. Using in vivo and in vitro experiments such as RT-qPCR, Western Blot, immunohistochemistry, and subcutaneous tumor formation in mice, we further validated the functions of key nucleotide metabolism genes in cell lines and animals. Nucleotide metabolism genes classified LUAD patients into two distinct subtypes with significant prognostic differences. The 'C1' subtype associated with active nucleotide metabolism pathways showed poorer prognosis and a more aggressive tumor phenotype. Furthermore, a nucleotide metabolism-related score (NMRS) calculated from the expression of 28 key genes effectively differentiated between patient outcomes and predicted associations with oncogenic pathways and immune responses. By integrating various immune infiltration algorithms, we delineated the associations between nucleotide metabolism signature genes and the tumor microenvironment, and characterized their distribution differences at the cellular level by analyzing single-cell sequencing dataset related to immunochemotherapy. Finally, we demonstrated the differential expression of the key nucleotide metabolism gene AUNIP acts as an oncogene to promote LUAD cell proliferation and is associated with tumor immune infiltration. The study underscores the pivotal role of nucleotide metabolism in LUAD progression and prognosis, highlighting the NMRS as a valuable biomarker for clinical outcomes and therapeutic responses. Specifically, AUNIP functions as a critical oncogene, offering a promising target for novel treatment strategies in LUAD.

A Study of the Relationship Between Driving and Health Based on Large-Scale Data Analysis Using PLSA and t-SNE

A Study of the Relationship Between Driving and Health Based on Large-Scale Data Analysis Using PLSA and t-SNE

Harnessing AI for Optimizing Formulation Components in Advanced Drug Delivery Systems: Analysis of Large-Scale Screening Data

Harnessing AI for Optimizing Formulation Components in Advanced Drug Delivery Systems: Analysis of Large-Scale Screening Data

Gender differences in research fields of bioeconomy and rural development-based on sustainable systems in Latin America and Africa regions.

Using bibliometric analysis of large-scale publication data is a simple approach to exploring gender-related trends, especially gender equality in academic publishing. The aim of this study is to investigate gender trends in the fields of bio-economy and rural development sciences in two under develop regions as Latin America and Africa. This study examines gender differences in these fields in order to: (1) recognize the contribution of female researchers in bioeconomy and rural development, (2) explore the relational structure of gender aspects in academic publications, (3) identify trends in female authorship in these scientific research fields over time, and finally (4) identify gender potentials for women to become more visible in these fields of study. To achieve these objectives, we used bibliometric tools to analyses 1891 publication records in bioeconomy and rural development. After cleaning the database of full names of authors of academic publications relevant to the field studies, we performed a series of statistical analyses in R and SPSS software, such as Lotkas distribution, network analysis, co-authorship analysis and spatial distribution of authors in the study. The results show that the number of male authors is almost three times higher than the number of female authors, suggesting that women are under-represented in the fields studied. Men occupy the most important position of authorship in scientific articles; publications with corresponding male authors were found in 1389 out of 1891 publications related to the bio-economy and rural development. In terms of geographical regions, publications with female authors were more prevalent in European and North American areas, with a small exception in some developing countries such as Argentina and South Africa. In terms of research networks, from the total number of authors evaluated, only 23% are female authors on the map of research influence. This indicates that there is a significant gap to be filled in the promotion of scholarly impact through the sharing of knowledge and expertise among authors.

Health and comfort oriented automatic generative design and optimization of residence space layout: An integrated data-driven and knowledge-based approach

During the initial design phase, space layout design plays a critical role in ensuring the ultimate performance of a residence. However, physical performance like energy consumption is usually prioritized, while health and comfort performance of residence is less explored. This study proposes an integrated data-driven and knowledge-based approach for automatic generative design of residence space layout focusing on health and comfort performance. First, explicit knowledge about key parameters of the layout design is obtained through quantitative analysis of large-scale health data. Second, with health and comfort as performance oriented, adjacency preference and noise score are introduced as objective functions and automatic generative design is carried out to obtain multiple layout alternatives. Finally, multi-objective deviation analysis is conducted to determine the optimal layout scheme. This study offers a practical and efficient tool through the whole design process, including pre-parameters analysis, during-process generation, and post-scheme evaluation, which contributes to improve decision-making.

Database profile of the Fukuoka-City Information Platform for Community-based Integrated Care.

The Fukuoka-City Information Platform for Community-based Integrated Care is an advanced big data platform that aggregates information on the health and medical services of Fukuoka citizens. Fukuoka City is engaged in a joint project with Kyushu University to promote policy making through a large-scale real-world data analysis. This paper describes the framework for this cooperative effort and the features of the analytical platform. Fukuoka City is the fifth most populous ordinance-designated city in Japan, with an estimated population of approximately 1.6 million. Under an agreement with Fukuoka City, Kyushu University was granted access to a portion of the city's anonymized healthcare database as secondary-use information. The database contains information on resident registration, health insurance claims, specific health checkups and health checkups for the older adults, specific health guidance, long-term care insurance data, and cancer screenings collected after fiscal year 2012. Each of these constituent datasets can be interlinked using anonymized hashed key variables, allowing individuals to be followed across databases and over time. The platform allows longitudinal investigation of the complex association between various aspects of healthcare, such as medical procedures, examinations, interviews, medical costs, long-term care certifications, and care costs. The platform can provide valuable public-health information because it is relatively large for a single database, and because it allows analysis of data across multiple domains and tracing of individuals over time.

HEaaN-STAT: a privacy-preserving statistical analysis toolkit for large-scale numerical, ordinal, and categorical data

Statistical analysis of largescale data is useful as it enables the extraction of a large amount of information, despite its simplicity. Therefore, fusing and analyzing data from different security domains is an attractive and promising approach, unless it jeopardizes the privacy of the data in any security domain. In this study, we proposed the HEaaN-STAT toolkit that can efficiently fuse data from different domains to enable largescale statistical analysis while protecting data privacy. Moreover, we proposed an efficient inverse operation and a table lookup function for Cheon-Kim-Kim-Song (CKKS) encrypted data, as well as a data encoding method for counting encrypted data. Based on this, we proposed a method for generating a contingency table with a large number of cases and k-percentile for largescale data that is hundreds to thousands of times faster than the method proposed by Lu et al. in NDSS' 17. The validity of the proposed toolkit was verified through practical use for business applications using real-world data.

Streamlined Algorithms for Large-Scale Matrix Computations in High-Performance Computing

In the realm of high-performance computing (HPC), large-scale matrix computations are fundamental to a myriad of scientific and engineering applications, from simulations in physics and engineering to data analysis in machine learning. This paper addresses the critical need for efficient and scalable algorithms capable of handling the growing computational demands of these applications. This study explores advanced algorithmic strategies that leverage the architecture of modern HPC systems, including parallelism, distributed computing, and optimized memory usage. The proposed algorithms are designed to minimize computational complexity and maximize performance, ensuring that largescale matrix operations can be executed swiftly and accurately. One of the key contributions of this work is the development of parallel algorithms that distribute the computational workload across multiple processors, significantly reducing execution time. By employing techniques such as data partitioning and workload balancing, the algorithms ensure optimal use of available computational resources, thus enhancing scalability. Furthermore, the paper introduces innovative approaches to memory management, which are crucial for handling large matrices that exceed the memory capacity of individual computing nodes. These approaches include efficient data storage formats and dynamic memory allocation strategies that reduce overhead and improve data access speeds. Another critical aspect of this research is the application of these algorithms to real-world problems. The paper presents case studies in fields such as climate modeling, structural analysis, and machine learning, demonstrating the practical benefits and performance gains achieved through the use of streamlined matrix computation algorithms. The results show substantial improvements in computation time and resource utilization, validating the effectiveness of the proposed methods. Additionally, the paper discusses the integration of these algorithms into existing HPC frameworks and libraries, making them accessible to a broader community of researchers and practitioners. By providing detailed implementation guidelines and performance benchmarks, the study ensures that the benefits of these advanced algorithms can be readily harnessed in various applications. The research also delves into the theoretical underpinnings of the proposed algorithms, offering a rigorous analysis of their computational complexity and scalability. In conclusion, this paper makes significant strides in addressing the challenges of large-scale matrix operations. Through the development of efficient, scalable, and practical algorithms, this research enhances the capability of HPC systems to tackle complex computational tasks, paving the way for advancements in various scientific and engineering domains. The proposed methods not only improve performance and resource utilization but also contribute to the broader goal of making high-performance computing more accessible and effective for large-scale data analysis and simulation.