Library Dataset Research Articles

Version [formula omitted] — [CACP: Classification Algorithms Comparison Pipeline]

We present the first major release of the Classification Algorithms Comparison Pipeline (CACP). The proposed software enables one to compare newly developed classification algorithms in Python with other classifiers to evaluate classification performance and ensure both outcomes’ reproducibility and statistical reliability. CACP simplifies and accelerates the entire classifier evaluation process considerably and helps prepare the professional documentation of the experiments conducted. The upgrade introduces enhancements to existing tools and adds new features: (1) - support for River machine learning library datasets in incremental learning, (2) - capability to include user-defined datasets, (3) - use of River classifiers for incremental learning, (4) - use of River metrics for incremental learning, (5) - flexibility to create user-defined metrics, (6) - record-by-record testing for incremental learning, (7) - enhanced summary of incremental testing results with dynamic visualization of the learning process, (8) - Graphical User Interface (GUI).

Associations of Maternal Nutritional Status and Supplementation with Fetal, Newborn, and Infant Outcomes in Low-Income and Middle-Income Settings: An Overview of Reviews

Background/Objectives: Despite advances in maternal nutritional knowledge, the effect of maternal diet, micronutrient status and undernutrition, and the effect of maternal supplementation on fetal, neonatal and infant outcomes still have gaps in the literature. This overview of reviews is intended to assess the available information on these issues and identify the main maternal nutritional factors associated with offspring outcomes in low- and middle-income countries as possible targets for public health interventions. Methods: The literature search was performed in Medline (PubMed) and Cochrane Library datasets in June 2024. Pre-specified outcomes in offspring were pooled using standard meta-analytical methods. Results: We found consistent evidence on the impact of maternal undernutrition indicated by low body mass index (BMI), mid-upper arm circumference (MUAC), and stature, but not of individual micronutrient status, on intrauterine-growth retardation, preterm birth, low birth weight, and small for gestational age, with research showing a possible effect of maternal undernutrition in later child nutritional status. Studies on micronutrient supplementation showed possible beneficial effects of iron, vitamin D, and multiple micronutrients on birthweight and/or decreasing small for gestational age, as well as a possible effect of calcium on preterm birth reduction. Interventions showing more consistent beneficial outcomes were balanced protein-energy and lipid base supplements, which demonstrated improved weight in newborns from supplemented mothers and a decreased risk of adverse neonatal outcomes. Conclusions: Further research is needed to identify the benefits and risks of maternal individual micronutrient supplementation on neonatal and further child outcomes.

The Data Analysis of Enterprise Operational Risk Prediction Under Machine Learning

In the digital age, the financial sector faces increasingly severe risk management challenges. Traditional methods often rely on historical data and statistical models, which struggle to cope with the high volatility of the market. These methods exhibit poor adaptability in rapidly changing financial markets and often fail to meet demands in terms of accuracy and reliability. To address these issues, this study proposes a law risk prediction model based on deep learning—the WBIF model. This model integrates Bidirectional Long Short-Term Memory (BiLSTM) and Fully Convolutional Networks (FCN) and employs the Whale Optimization Algorithm (WOA) for parameter optimization. Experimental results show that compared to traditional models, the WBIF model reduces the Mean Absolute Error (MAE) by 51.73% on the UCI machine learning library dataset and improves accuracy by 12% on the Kaggle credit card fraud detection dataset.

Contrast-enhanced CT scan (CECT) for the detection of hollow viscus and mesenteric injuries in blunt trauma - an updated systematic review of the literature and meta-analysis of diagnostic test accuracy.

Despite improved awareness of blunt traumatic hollow viscus and mesenteric injuries (THVMI), the accuracy of contrast-enhanced CT (CECT) varies considerably among studies. This systematic review and meta-analysis of test accuracy aims to explore the diagnostic performance of CECT in detecting THVMI in blunt trauma. The study was conducted according to the Cochrane recommendations searching the PubMed, Scopus, and Cochrane Library datasets from 2000 to 8 September 2023 (PROSPERO ID: CRD42023473041). Surgical exploration, autopsy, and discharge from the hospital after monitoring were set as reference standard. To explore the diagnostic accuracy of CECT in detecting THVMI hierarchical models were developed. The risk of bias in individual studies was assessed with the QUADAS-2 tool. Sensitivity analysis was conducted to detect sources of heterogeneity. Twelve studies, for a total of 4537 patients, were deemed eligible. After identification of outliers and sensitivity analysis, the pooled sensitivity, specificity, positive and negative likelihood ratios, and diagnostic odds ratio were 0.85 (95% CI: 0.69-0.93), 0.94 (95% CI: 0.8-0.98), 14.65 (95% CI: 4.22-50.85), 0.16 (95% CI: 0.07-0.34), 92.3 (95% CI: 29.75-286.34), respectively. The Area under the HSROC curve was 0.95 (95% CI: 0.92-0.96). Meta-regression analysis identified the year of publication as a covariate significantly associated with heterogeneity. A high risk of bias was detected in the "patient selection" domains. CECT has a fundamental role in identifying THVMI with high specificity but suboptimal sensitivity. Clinical criteria are still of paramount importance, especially in cases of ambiguous initial CECT images.

Data-driven ergonomic assessment of construction workers

Automating ergonomic assessment improves both objectivity and cost-effectiveness. However, existing ergonomic scales remain susceptible to inaccuracy and insensitivity when assessing diversified construction activities. This susceptibility stems from the unreliable ranges, sharp boundaries, and oversimplified binary rules within conventional joint-level assessment rules, coupled with the restricted transformation rules to integer input/output. This paper presents a data-driven ergonomic assessment method grounded in statistical data from Construction Motion Data Library (CML) dataset, comprising: 1) Developing joint-level scoring models to primarily improve accuracy through leveraging Heuristics Gaussian Cloud Transformation (H-GCT), and 2) Designing fuzzy inference mechanism to enhance sensitivity through retaining risk information during transformation across levels. Subsequently, a three-step validation confirms accurate risk identification, sensitive risk feedback, and positive correlation with previous methods, contributing to occupational health and safety management. Future research could enhance the method by expanding the pose dataset, recruiting more participants, and refining the 3D pose estimation.

REAL-Colon: A dataset for developing real-world AI applications in colonoscopy

Detection and diagnosis of colon polyps are key to preventing colorectal cancer. Recent evidence suggests that AI-based computer-aided detection (CADe) and computer-aided diagnosis (CADx) systems can enhance endoscopists' performance and boost colonoscopy effectiveness. However, most available public datasets primarily consist of still images or video clips, often at a down-sampled resolution, and do not accurately represent real-world colonoscopy procedures. We introduce the REAL-Colon (Real-world multi-center Endoscopy Annotated video Library) dataset: a compilation of 2.7 M native video frames from sixty full-resolution, real-world colonoscopy recordings across multiple centers. The dataset contains 350k bounding-box annotations, each created under the supervision of expert gastroenterologists. Comprehensive patient clinical data, colonoscopy acquisition information, and polyp histopathological information are also included in each video. With its unprecedented size, quality, and heterogeneity, the REAL-Colon dataset is a unique resource for researchers and developers aiming to advance AI research in colonoscopy. Its openness and transparency facilitate rigorous and reproducible research, fostering the development and benchmarking of more accurate and reliable colonoscopy-related algorithms and models.

Incorporating Synthetic Accessibility in Drug Design: Predicting Reaction Yields of Suzuki Cross-Couplings by Leveraging AbbVie's 15-Year Parallel Library Data Set.

Despite the increased use of computational tools to supplement medicinal chemists' expertise and intuition in drug design, predicting synthetic yields in medicinal chemistry endeavors remains an unsolved challenge. Existing design workflows could profoundly benefit from reaction yield prediction, as precious material waste could be reduced, and a greater number of relevant compounds could be delivered to advance the design, make, test, analyze (DMTA) cycle. In this work, we detail the evaluation of AbbVie's medicinal chemistry library data set to build machine learning models for the prediction of Suzuki coupling reaction yields. The combination of density functional theory (DFT)-derived features and Morgan fingerprints was identified to perform better than one-hot encoded baseline modeling, furnishing encouraging results. Overall, we observe modest generalization to unseen reactant structures within the 15-year retrospective library data set. Additionally, we compare predictions made by the model to those made by expert medicinal chemists, finding that the model can often predict both reaction success and reaction yields with greater accuracy. Finally, we demonstrate the application of this approach to suggest structurally and electronically similar building blocks to replace those predicted or observed to be unsuccessful prior to or after synthesis, respectively. The yield prediction model was used to select similar monomers predicted to have higher yields, resulting in greater synthesis efficiency of relevant drug-like molecules.

Influence government, platform power and the patchwork profile: Exploring the appropriation of targeted advertising infrastructures for government behaviour change campaigns

The targeted digital advertising infrastructures on which the business models of the social media platform economy rest have been the subject of significant academic and political interest. In this paper, we explore and theorise the appropriation of these infrastructures — designed for commercial and political advertising — by the state. In the U.K., public sector bodies have begun to repurpose the surveillance and messaging capacities of these social media platforms, along with the influencer economy, to deliver targeted behaviour change campaigns to achieve public policy goals. We explore how frameworks of behavioural government have aligned with Internet platforms’ extensive infrastructures and the commercial ecologies of professionalised strategic marketing. We map the current extent of these practices in the U.K. through case studies and empirical research in Meta’s Ad Library dataset. Although the networks of power and discourse within the ad infrastructure are indeed acting to shape the capacities of the state to engage in online influence, public bodies are mobilising their own substantial material networks of power and data to re-appropriate them to their own ends. Partly as a result of attempts by Meta to restrict the targeting of protected characteristics, we observe state communications campaigns building up what we term patchwork profiles of minute behavioural, demographic, and location-based categories in order to construct and reach particular groups of subjects. However, rather than a clear vision of a ‘cybernetic society’ of reactive information control, we instead find a heterogeneous and piecemeal landscape of different modes of power.

MARS an improved de novo peptide candidate selection method for non-canonical antigen target discovery in cancer

Understanding the nature and extent of non-canonical human leukocyte antigen (HLA) presentation in tumour cells is a priority for target antigen discovery for the development of next generation immunotherapies in cancer. We here employ a de novo mass spectrometric sequencing approach with a refined, MHC-centric analysis strategy to detect non-canonical MHC-associated peptides specific to cancer without any prior knowledge of the target sequence from genomic or RNA sequencing data. Our strategy integrates MHC binding rank, Average local confidence scores, and peptide Retention time prediction for improved de novo candidate Selection; culminating in the machine learning model MARS. We benchmark our model on a large synthetic peptide library dataset and reanalysis of a published dataset of high-quality non-canonical MHC-associated peptide identifications in human cancer. We achieve almost 2-fold improvement for high quality spectral assignments in comparison to de novo sequencing alone with an estimated accuracy of above 85.7% when integrated with a stepwise peptide sequence mapping strategy. Finally, we utilize MARS to detect and validate lncRNA-derived peptides in human cervical tumour resections, demonstrating its suitability to discover novel, immunogenic, non-canonical peptide sequences in primary tumour tissue.

Using explainable machine learning to uncover the kinase-substrate interaction landscape.

Phosphorylation, a post-translational modification regulated by protein kinase enzymes, plays an essential role in almost all cellular processes. Understanding how each of the nearly 500 human protein kinases selectively phosphorylates their substrates is a foundational challenge in bioinformatics and cell signaling. Although deep learning models have been a popular means to predict kinase-substrate relationships, existing models often lack interpretability and are trained on datasets skewed toward a subset of well-studied kinases. Here we leverage recent peptide library datasets generated to determine substrate specificity profiles of 300 serine/threonine kinases to develop an explainable Transformer model for kinase-peptide interaction prediction. The model, trained solely on primary sequences, achieved state-of-the-art performance. Its unique multitask learning paradigm built within the model enables predictions on virtually any kinase-peptide pair, including predictions on 139 kinases not used in peptide library screens. Furthermore, we employed explainable machine learning methods to elucidate the model's inner workings. Through analysis of learned embeddings at different training stages, we demonstrate that the model employs a unique strategy of substrate prediction considering both substrate motif patterns and kinase evolutionary features. SHapley Additive exPlanation (SHAP) analysis reveals key specificity determining residues in the peptide sequence. Finally, we provide a web interface for predicting kinase-substrate associations for user-defined sequences and a resource for visualizing the learned kinase-substrate associations. All code and data are available at https://github.com/esbgkannan/Phosformer-ST. Web server is available at https://phosformer.netlify.app.

Reusing Peptide Spectral Reference Libraries to Discover Putative Plasma Biomarkers of Response to Cancer Chemotherapy.

Biofluids such as blood plasma are rich reservoirs of potential biomarkers for disease diagnosis, prognosis, and prediction of treatment response. However, mass spectrometry analysis of circulating plasma proteins remains challenging. The introduction of data-independent acquisition mass spectrometry (DIA-MS) is an important step toward addressing detection of less abundant plasma proteins. Numerous plasma peptide MS/MS spectral library datasets produced from extensive plasma fractionation are accessible from public archives, and these can be repurposed as spectral reference libraries to increase the depth of proteomic analysis when DIA-MS is used. Here we describe the workflow that relies on reusing the existing spectral reference libraries by populating them with locally obtained peptide MS/MS data acquired by DIA-MS. This approach was demonstrated effectively to identify putative plasma biomarkers of response to neoadjuvant chemotherapy in the setting of pancreatic ductal adenocarcinoma (PDAC) (O'Rourke et al., J Proteomics 231:103998, 2021).

A two‐layer approach for solving robust decentralized multiproject scheduling problem with multi‐skilled staff

AbstractUncertain activity durations and multi‐skilled staff allocation are two practical factors that are difficult to cope with in real project management. This paper studies a new decentralized resource‐constrained multiproject scheduling problem with uncertain activity durations and multi‐skilled staff allocation. The robust project scheduling is employed to tackle uncertainty. We name it the robust decentralized resource‐constrained multiproject scheduling problem with multi‐skilled staff (RDMPSP‐MS). To formulate this problem, a two‐layer model containing local scheduling and global coordination decision‐making is established based on the multiagent system. To solve this model, a two‐layer approach (TLA) is developed. In the TLA, a starting time criticality heuristic method is adopted to handle the local scheduling problem, and a global resource allocation heuristic algorithm is designed to allocate multi‐skilled staff. The computational experiments are conducted on the adapted Multi‐Project Scheduling Problem LIBrary (MPSPLIB) data set. The results show that the proposed TLA outperforms seven heuristic algorithms based on priority rules regarding solution quality and efficiency on most problem subsets. It is further verified that the robustness of the baseline schedule deteriorates with the increase in problem size and activity duration variability level. Additionally, the results of this research provide practical guidance for managers who expect to determine reasonable project deadlines in a decentralized multiproject environment.

Multiple-input streams attention (MISA) network for skeleton-based construction workers' action recognition using body-segment representation strategies

With the rapid growth of deep learning algorithms, graph convolutional networks (GCNs) have become a common choice for skeleton-based human action recognition, boasting impressive performance. However, existing GCN-based models often rely on physical human body connections, which may not suit complex construction tasks involving various body parts and hand movements. To address this concern, the human body is modeled in this paper through topological graphs at varying levels, designed based on body-segment strategies. A multiple-input streams attention (MISA) network is introduced, incorporating GCN and temporal convolutional network (TCN) components to enhance the body-structure topology graph of GCNs with more comprehensive input graphs. Additionally, two-modality motion data and three attention blocks are integrated to capture more discerning features. Finally, experimental results using the Construction Motion Library (CML) dataset demonstrated the superiority of the developed method, reaching approximately 84.94% recognition accuracy.

DNA methylation profiling to determine the primary sites of metastatic cancers using formalin-fixed paraffin-embedded tissues

Identifying the primary site of metastatic cancer is critical to guiding the subsequent treatment. Approximately 3–9% of metastatic patients are diagnosed with cancer of unknown primary sites (CUP) even after a comprehensive diagnostic workup. However, a widely accepted molecular test is still not available. Here, we report a method that applies formalin-fixed, paraffin-embedded tissues to construct reduced representation bisulfite sequencing libraries (FFPE-RRBS). We then generate and systematically evaluate 28 molecular classifiers, built on four DNA methylation scoring methods and seven machine learning approaches, using the RRBS library dataset of 498 fresh-frozen tumor tissues from primary cancer patients. Among these classifiers, the beta value-based linear support vector (BELIVE) performs the best, achieving overall accuracies of 81-93% for identifying the primary sites in 215 metastatic patients using top-k predictions (k = 1, 2, 3). Coincidentally, BELIVE also successfully predicts the tissue of origin in 81-93% of CUP patients (n = 68).

Comparison of clinical outcomes of arthroscopic rotator cuff repair utilizing suture-bridge procedures with or without medial knots: a meta-analysis

PurposeThis investigation aimed to compare the medical efficacy of the knotted and knotless suture-bridge procedures in rotator cuff repair.MethodsThe Pubmed, Embase, and Cochrane Library datasets were searched for all available publications comparing the medical results of arthroscopic rotator cuff repairs utilizing knotted or knotless suture-bridge procedures. Two researchers utilized Newcastle-Ottawa Scale and Cochrane risk-of-bias tool to evaluate the included studies. Employing Revman 5.3 software, meta-analysis was conducted following the PRISMA reporting guideline.ResultsEleven investigations with 1083 patients were considered suitable for the final meta-analysis. 522 individuals were assigned to the knotted group, whereas 561 were assigned to the knotless group. No statistical difference was found between the knotted and knotless groups, regarding VAS score (WMD, 0.17; 95% CI, − 0.10 to 0.44; P = 0.21); Constant score (WMD, -1.50; 95% CI, − 3.52 to 0.52; P = 0.14); American Shoulder and Elbow Surgeons Shoulder (WMD, -2.02; 95% CI, − 4.53 to 0.49; P = 0.11); University of California Los Angeles score (WMD, -0.13; 95% CI, − 0.89 to 0.63; P = 0.73); ROM of flexion (WMD, 1.57; 95% CI, − 2.11 to 5.60; P = 0.37), abduction (WMD, 1.08; 95% CI, − 4.53 to 6.70; P = 0.71) and external rotation (WMD, 1.90; 95% CI, − 1.36 to 5.16; P = 0.25); re-tear rate (OR, 0.74; 95% CI, 0.50 to 1.08; P = 0.12), and medical complications (OR, 0.90; 95% CI, 0.37 to 2.20; P = 0.82).ConclusionFor arthroscopic rotator cuff repairs, there were no statistical differences in medical results among knotted and knotless suture-bridge procedures. Overall, both techniques showed excellent clinical outcomes and could be safely utilized to treat rotator cuff injuries.

Evaluating explainability for graph neural networks

As explanations are increasingly used to understand the behavior of graph neural networks (GNNs), evaluating the quality and reliability of GNN explanations is crucial. However, assessing the quality of GNN explanations is challenging as existing graph datasets have no or unreliable ground-truth explanations. Here, we introduce a synthetic graph data generator, ShapeGGen, which can generate a variety of benchmark datasets (e.g., varying graph sizes, degree distributions, homophilic vs. heterophilic graphs) accompanied by ground-truth explanations. The flexibility to generate diverse synthetic datasets and corresponding ground-truth explanations allows ShapeGGen to mimic the data in various real-world areas. We include ShapeGGen and several real-world graph datasets in a graph explainability library, GraphXAI. In addition to synthetic and real-world graph datasets with ground-truth explanations, GraphXAI provides data loaders, data processing functions, visualizers, GNN model implementations, and evaluation metrics to benchmark GNN explainability methods.

Point mutations of homologs as an adaptive solution to the gene loss

Gene loss is common and influences genome evolution trajectories. Multiple adaptive strategies to compensate for gene loss have been observed, including copy number gain of paralogous genes and mutations in genes of the same pathway. By using the Ubl-specific protease 2 (ULP2) eviction model, we identify compensatory mutations in the homologous gene ULP1 by laboratory evolution and find that these mutations are capable of rescuing defects caused by the loss of ULP2. Furthermore, bioinformatics analysis of genomes of yeast gene knockout library and natural yeast isolate datasets suggests that point mutations of a homologous gene might be an additional mechanism to compensate for gene loss.

Rock image classification using deep residual neural network with transfer learning

Rock image classification is a significant part of geological research. Compared with traditional image classification methods, rock image classification methods based on deep learning models have the great advantage in terms of automatic image features extraction. However, the rock classification accuracies of existing deep learning models are unsatisfied due to the weak feature extraction ability of the network model. In this study, a deep residual neural network (ResNet) model with the transfer learning method is proposed to establish the corresponding rock automatic classification model for seven kinds of rock images. ResNet34 introduces the residual structure to make it have an excellent effect in the field of image classification, which extracts high-quality rock image features and avoids information loss. The transfer learning method abstracts the deep features from the shallow features, and better express the rock texture features for classification in the case of fewer rock images. To improve the generalization of the model, a total of 3,82,536 rock images were generated for training via image slicing and data augmentation. The network parameters trained on the Texture Library dataset which contains 47 types of texture images and reflect the characteristics of rocks are used for transfer learning. This pre-trained weight is loaded when training the ResNet34 model with the rock dataset. Then the model parameters are fine-tuned to transfer the model to the rock classification problem. The experimental results show that the accuracy of the model without transfer learning reached 88.1%, while the model using transfer learning achieved an accuracy of 99.1%. Aiming at geological engineering field investigation, this paper studies the embedded deployment application of the rock classification network. The proposed rock classification network model is transplanted to an embedded platform. By designing a rock classification system, the off-line rock classification is realized, which provides a new solution for the rock classification problem in the geological survey. The deep residual neural network and transfer learning method used in this paper can automatically classify rock features without manually extracting. These methods reduce the influence of subjective factors and make the rock classification process more automatic and intelligent.

Machine Learning Screens Potential Drugs Targeting a Prognostic Gene Signature Associated With Proliferation in Hepatocellular Carcinoma.

Background: This study aimed to screen potential drugs targeting a new prognostic gene signature associated with proliferation in hepatocellular carcinoma (HCC). Methods: CRISPR Library and TCGA datasets were used to explore differentially expressed genes (DEGs) related to the proliferation of HCC cells. Differential gene expression analysis, univariate COX regression analysis, random forest algorithm and multiple combinatorial screening were used to construct a prognostic gene signature. Then the predictive power of the gene signature was validated in the TCGA and ICGC datasets. Furthermore, potential drugs targeting this gene signature were screened. Results: A total of 640 DEGs related to HCC proliferation were identified. Using univariate Cox analysis and random forest algorithm, 10 hub genes were screened. Subsequently, using multiplex combinatorial screening, five hub genes (FARSB, NOP58, CCT4, DHX37 and YARS) were identified. Taking the median risk score as a cutoff value, HCC patients were divided into high- and low-risk groups. Kaplan-Meier analysis performed in the training set showed that the overall survival of the high-risk group was worse than that of the low-risk group (p < 0.001). The ROC curve showed a good predictive efficiency of the risk score (AUC > 0.699). The risk score was related to gene mutation, cancer cell stemness and immune function changes. Prediction of immunotherapy suggetsted the IC50s of immune checkpoint inhibitors including A-443654, ABT-888, AG-014699, ATRA, AUY-922, and AZ-628 in the high-risk group were lower than those in the low-risk group, while the IC50s of AMG-706, A-770041, AICAR, AKT inhibitor VIII, Axitinib, and AZD-0530 in the high-risk group were higher than those in the low-risk group. Drug sensitivity analysis indicated that FARSB was positively correlated with Hydroxyurea, Vorinostat, Nelarabine, and Lomustine, while negatively correlated with JNJ-42756493. DHX37 was positively correlated with Raltitrexed, Cytarabine, Cisplatin, Tiotepa, and Triethylene Melamine. YARS was positively correlated with Axitinib, Fluphenazine and Megestrol acetate. NOP58 was positively correlated with Vorinostat and 6-thioguanine. CCT4 was positively correlated with Nerabine. Conclusion: The five-gene signature associated with proliferation can be used for survival prediction and risk stratification for HCC patients. Potential drugs targeting this gene signature deserve further attention in the treatment of HCC.

The Brazilian Program of soil analysis via spectroscopy (ProBASE): Combining spectroscopy and wet laboratories to understand new technologies

Spectroscopy has been extensively used in soil analysis. However, users, such as commercial laboratories, still do not know the application, potential, and limitations of this technique. This paper aims to present a given course going from theory to practice for end-users regarding spectroscopy technique. The paper combines the use of soil spectroscopy and wet laboratories to perform a scientific-teaching dynamic using their own data, allowing for a better understanding of the technique. The course is denominated ProBASE (The Brazilian Program of Soil Analysis via Spectroscopy). Soil samples from 35 laboratories were sent (34 from Brazil and 1 from Paraguay) to a central spectroscopy laboratory. Samples were measured for visible-near-short-wave-infrared, Vis-NIR (400–2500 nm), and mid-infrared, MIR (3000–25000 nm) ranges and by portable X-ray Fluorescence (pXRF) sensor. We also used the Brazilian Soil Spectral Library (BSSL) dataset (Vis-NIR). We performed three different population models with Vis-NIR as follows: a) using the dataset of each laboratory (Local), b) using the entire ProBASE-dataset (Regional), and c) using the BSSL (Country). Afterward, we developed spectral models using the other spectral ranges for comparison. We also used a qualitative approach to detect errors from the wet laboratory analyses using Vis-NIR data and evaluated their impact on spectral modeling. The models that used Local samples had the best performance, with R2 in validation reaching up to 0.93 for clay, 0.92 for sand, 0.86 for P, 0.82 for pH, 0.81 for organic matter (OM), 0.75 for Ca2+, 0.72 for cation exchange capacity (CEC), 0.71 for aluminum saturation, 0.71 for Al3+, 0.70 for Mg2+, 0.64 for base saturation (V%), and 0.56 for K+. However, the base saturation presented greater variation from good to poor results. For the comparison dataset, the results can be summarized as follows: a) pXRF was better for P, Ca2+ B, V% and Mn; b) MIR was better for clay, sand, OM, pH, Mg2+, CEC and Mn; c) Vis-NIR was better for H + Al; d) the three spectral ranges had good performance for OM, sand, silt and clay. In addition, our findings indicate that all spectral ranges are useful for a wet laboratory, where each model has advantages and limitations, but they can be used complementary to each other. Spectroscopy can detect inconsistencies of the wet laboratory analyses, affecting thus the quality of the results. The commercial laboratory community viewed the techniques positively. The results indicate the viability to create a hybrid laboratory, combining both wet and dry (soil spectroscopy) chemistry.