Selection Method Research Articles

Network Analysis of Enhancer-Promoter Interactions Highlights Cell-Type-Specific Mechanisms of Transcriptional Regulation Variation.

Gene expression is orchestrated by a complex array of gene regulatory elements that govern transcription in a cell-type-specific manner. Though previously studied, the ability to utilize regulatory elements to identify disrupting variants remains largely elusive. To identify important factors within these regions, we generated enhancer-promoter interaction (EPI) networks and investigated the presence of disease-associated variants that fall within these regions. Our study analyzed six neuronal cell types across neural differentiation, allowing us to examine closely related cell types and across differentiation stages. Our results expand upon previous findings of cell-type specificity of enhancer, promoter, and transcription factor binding sites. Notably, we find that regulatory regions within EPI networks can identify the enrichment of variants associated with neuropsychiatric disorders within specific cell types and network sub-structures. This enrichment within sub-structures can allow for a better understanding of potential mechanisms by which variants may disrupt transcription. Together, our findings suggest that EPIs can be leveraged to better understand cell-type-specific regulatory architecture and used as a selection method for disease-associated variants to be tested in future functional assays. Combined with these future functional characterization assays, EPIs can be used to better identify and characterize regulatory variants' effects on such networks and model their mechanisms of gene regulation disruption across different disorders. Such findings can be applied in practical settings, such as diagnostic tools and drug development.

Monitoring aboveground organs biomass of wheat and maize: A novel model combining ensemble learning and allometric theory

Accurate monitoring of crop organ biomass facilitates optimizing agronomic strategies to maximize yield or economic benefit. Unmanned aerial vehicle (UAV) is extensively employed for aboveground biomass (AGB) monitoring at the farm scale, but previous studies have mostly concentrated on total AGB rather than individual organ biomass. Furthermore, film-mulched crops, a widely used cropping pattern in northwest China, have received less attention for AGB monitoring. We aim to develop a novel model to precisely estimate the AGB of leaf (AGBLeaf), stem (AGBStem), and reproductive organs (AGBR) by UAV for film-mulched wheat and maize. The maize-wheat rotation field experiments with treatments of five nitrogen application amounts and three planting densities were conducted from 2021 to 2023, respectively. Firstly, we constructed allometric models at jointing, heading (tasseling), and grain filling stages by ground sampling data in 2021–2022. Next, the input feature set was obtained by feature selection methods (Lasso and Boruta) using UAV image data, and three traditional methods (partial least squares, ridge regression, and support vector machine) and three ensemble learning models (random forest, extreme gradient boosting, and local cascade ensemble (LCE)) were trained for AGBLeaf inversion based on the physically-based PROSAIL model simulation dataset. Finally, the optimal AGBLeaf inversion hybrid model was coupled with the allometric model to estimate the AGBStem and AGBR in 2022–2023. The results indicated that both wheat and maize organ biomass conformed to the allometric pattern. While feature selection helped reduce computation and complexity, but didn’t improve monitoring accuracy. The normalized root mean square error (NRMSE) of the optimal hybrid model (PROSAIL + Boruta + LCE) on the measured wheat and maize AGBLeaf datasets were 12.72 %–24.93 % and 19.65 %–25.16 %, respectively. After coupling the allometric model, the coefficient of determination (R2) of wheat and maize AGBStem were 0.64–0.85 and 0.63–0.68, and the NRMSE were 15.05 %–25.28 % and 24.10 %–27.06 %, respectively; and the corresponding R2 of AGBR was 0.67–0.76 and 0.72, and the NRMSE were 16.81 %–22.12 % and 21.66 %, for wheat and maize, respectively. Overall, the novel model performed well in film-mulched wheat and maize, providing a cost-effective approach for organ biomass monitoring. In the future, further validation of the model’s transferability is necessary to increase the potential for generalization in production practice.

The spike-and-slab quantile LASSO for robust variable selection in cancer genomics studies.

Data irregularity in cancer genomics studies has been widely observed in the form of outliers and heavy-tailed distributions in the complex traits. In the past decade, robust variable selection methods have emerged as powerful alternatives to the nonrobust ones to identify important genes associated with heterogeneous disease traits and build superior predictive models. In this study, to keep the remarkable features of the quantile LASSO and fully Bayesian regularized quantile regression while overcoming their disadvantage in the analysis of high-dimensional genomics data, we propose the spike-and-slab quantile LASSO through a fully Bayesian spike-and-slab formulation under the robust likelihood by adopting the asymmetric Laplace distribution (ALD). The proposed robust method has inherited the prominent properties of selective shrinkage and self-adaptivity to the sparsity pattern from the spike-and-slab LASSO (Roc̆ková and George, J Am Stat Associat, 2018, 113(521): 431-444). Furthermore, the spike-and-slab quantile LASSO has a computational advantage to locate the posterior modes via soft-thresholding rule guided Expectation-Maximization (EM) steps in the coordinate descent framework, a phenomenon rarely observed for robust regularization with nondifferentiable loss functions. We have conducted comprehensive simulation studies with a variety of heavy-tailed errors in both homogeneous and heterogeneous model settings to demonstrate the superiority of the spike-and-slab quantile LASSO over its competing methods. The advantage of the proposed method has been further demonstrated in case studies of the lung adenocarcinomas (LUAD) and skin cutaneous melanoma (SKCM) data from The Cancer Genome Atlas (TCGA).

High Spatiotemporal Resolution Biomimetic Thermoreceptors Realizing by Jointless p‐n Integration Thermoelectric Composites

AbstractHuman body temperature is a critical physiological indicator, reflecting metabolism and temperature regulation. Biomimetic temperature sensors with high temporal and spatial resolution are highly desired to emulate human skin's temperature perception capabilities. Here, a selective filtration method is proposed to fabricate a jointless p‐n thermoelectric module, integrating p‐type and n‐type materials with minimal interfacial barriers. The structure enables rapid response, minimal signal variation, and a robust linear relationship between open circuit voltage and temperature difference, making it suitable as a biomimetic thermoreceptor for enhancing humanoid robots with intelligent sensing capabilities. A smart thermoelectric sensing system with a high temporal and spatial resolution is developed using with jointless p‐n module as the sensing element. In temporal thermal sensing, the sensing system can timely and effectively identify the heat source, with a fast response time of <0.1 s. In addition, in spatial thermal sensing, the system reliably detects thermal stimuli within discrete regions of each measuring area of <1 × 1 cm2. This study provides an effective strategy for temperature sensors in thermoelectric sensing and biomimetic thermoreceptor applications.

A machine learning model based on CHAT-23 for early screening of autism in Chinese children.

Autism spectrum disorder (ASD) is a neurodevelopmental condition that significantly impacts the mental, emotional, and social development of children. Early screening for ASD typically involves the use of a series of questionnaires. With answers to these questionnaires, healthcare professionals can identify whether a child is at risk for developing ASD and refer them for further evaluation and diagnosis. CHAT-23 is an effective and widely used screening test in China for the early screening of ASD, which contains 23 different kinds of questions. We have collected clinical data from Wuxi, China. All the questions of CHAT-23 are regarded as different kinds of features for building machine learning models. We introduce machine learning methods into ASD screening, using the Max-Relevance and Min-Redundancy (mRMR) feature selection method to analyze the most important questions among all 23 from the collected CHAT-23 questionnaires. Seven mainstream supervised machine learning models were built and experiments were conducted. Among the seven supervised machine learning models evaluated, the best-performing model achieved a sensitivity of 0.909 and a specificity of 0.922 when the number of features was reduced to 9. This demonstrates the model's ability to accurately identify children for ASD with high precision, even with a more concise set of features. Our study focuses on the health of Chinese children, introducing machine learning methods to provide more accurate and effective early screening tests for autism. This approach not only enhances the early detection of ASD but also helps in refining the CHAT-23 questionnaire by identifying the most relevant questions for the diagnosis process.

Uncertainty quantification in high-dimensional linear models incorporating graphical structures with applications to gene set analysis.

The functions of genes in networks are typically correlated due to their functional connectivity. Variable selection methods have been developed to select important genes associated with a trait while incorporating network graphical information. However, no method has been proposed to quantify the uncertainty of individual genes under such settings. In this paper, we construct confidence intervals (CIs) and provide P-values for parameters of a high-dimensional linear model incorporating graphical structures where the number of variables p diverges with the number of observations. For combining the graphical information, we propose a graph-constrained desparsified LASSO (least absolute shrinkage and selection operator) (GCDL) estimator, which reduces dramatically the influence of high correlation of predictors and enjoys the advantage of faster computation and higher accuracy compared with the desparsified LASSO. Theoretical results show that the GCDL estimator achieves asymptotic normality. The asymptotic property of the uniform convergence is established, with which an explicit expression of the uniform CI can be derived. Extensive numerical results indicate that the GCDL estimator and its (uniform) CI perform well even when predictors are highly correlated. An R package implementing the proposed method is available at https://github.com/XiaoZhangryy/gcdl.

Exploring alternative polymer materials for joint liners: a software-guided material selection

Abstract This study explores the alternative polymer materials and selection process for joint implant liners, focusing on applying CES Selector software to identify suitable polymer materials. CES Selector provides an easy-to-use interface. It offers multiple selection methods, including boundary values and property constraints. Seven materials were excluded from the analysis, resulting in 19 potential candidates, including unconventional options like EVOH, PCTA, PESU, PI, PPA, PPC, PPSU, and PSU. The materials underwent evaluation based on key criteria, including tensile strength, Young's modulus, compressive strength, fatigue strength, and fracture toughness. Overall, TPU exhibited a remarkable combination of high mechanical strength and adaptable Young's modulus, making it a top contender. However, in other evaluation criteria, PI surpassed TPU, solidifying its potential as a superior choice. This systematic approach provides valuable insights for engineers and designers seeking innovative materials for joint implant liners. The study results broaden the range of materials that can be used in implant manufacturing, providing potential alternatives that offer better long-term durability and performance.

A comparison of climate drivers’ impacts on silage maize yield shock in Germany

AbstractExtreme weather events have become more frequent and severe with ongoing climate change, with a huge implication for the agricultural sector and detrimental effects on crop yield. In this study, we compare several combinations of climate indices and utilized the Least Absolute Shrinkage and Selection Operator (LASSO) to explain the probabilities of substantial drops in silage maize yield (here defined as “yield shock” by using a 15th percentile as threshold) in Germany between 1999 and 2020. We compare the variable importance and the predictability skill of six combinations of climate indices using the Matthews Correlation Coefficient (MCC). Finally, we delve into year-to-year predictions by comparing them against the historical series and examining the variables contributing to high and low predicted yield shock probabilities. We find that cold conditions during April and hot and/or dry conditions during July increase the chance of silage maize yield shock. Moreover, a combination of simple variables (e.g. total precipitation) and complex variables (e.g. cumulative cold under cold nights) enhances predictive accuracy. Lastly, we find that the years with higher predicted yield shock probabilities are characterized mainly by relatively hotter and drier conditions during July compared to years with lower yield shock probabilities. Our findings enhance our understanding of how weather impacts maize crop yield shocks and underscore the importance of considering complex variables and using effective selection methods, particularly when addressing climate-related events.

Advanced Sorbents for Separation of Metal Ions

Effective, sustainable, and selective methods for recovering or removing metals from various media, such as mining leachates, recycling waste, industrial effluents, and natural water, are necessary due to the increasing demand for metals and stringent environmental constraints [...]

Ultrasound based radiomics model for assessment of placental function in pregnancies with preeclampsia

The goal of our research is to elucidate and better assess placental function in rats with preeclampsia through an innovative application of ultrasound-based radiomics. Using a rat model induced with L-NAME, we carefully investigated placental dysfunction via microstructural analysis and immunoprotein level assessment. Employing the Boruta feature selection method on ultrasound images facilitated the identification of crucial features, consequently enabling the development of a robust model for classifying placental dysfunction. Our study included 12 pregnant rats, and thorough placental evaluations were conducted on 160 fetal rats. Distinct alterations in placental microstructure and angiogenic factor expression were evident in rats with preeclampsia. Leveraging high-throughput mining of quantitative image features, we extracted 558 radiomic features, which were subsequently used to construct an impressive evaluation model with an area under the receiver operating curve (AUC) of 0.95. This model also exhibited a remarkable sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of 88.7%, 91.5%, 90.2%, 90.4%, and 90.0%, respectively. Our findings highlight the ability of ultrasound-based radiomics to detect abnormal placental features, demonstrating its potential for evaluating both normative and impaired placental function with high precision and reliability.

Simultaneous detection of enrofloxacin and chloramphenicol antibiotics using surface-enhanced Raman spectroscopy combined with triangular silver nanoplates

The emergence of antibiotic residues in the environment and food requires the development of sensitive and selective detection methods for monitoring these contaminants. In fact, many different types of antibiotics will be used simultaneously to prevent diseases in livestock and aquatic animals, causing antibiotic residues in food and the environment. Rapid and accurate simultaneous analysis of multiple antibiotics at low concentrations remains a major challenge in in the field of rapid detection. This study presents a sensitive and selective method for the simultaneous detection of two widely used antibiotics, enrofloxacin and chloramphenicol, by combining surface-enhanced Raman spectroscopy (SERS) with triangular silver nanoplates. Triangular silver nanoplates (TAgNPls) in the form of colloid were simply synthesized by chemical reduction method and then used as an effective SERS substrate, enhancing the Raman signals of the target antibiotics and enabling the detection of these antibiotics at trace levels. It was found that with a SERS substrate assembled from the above TAgNPls, the antibiotics enrofloxacin and chloramphenicol could be detected with detection limits of 2.6 µg/L and 4.3 µg/L, respectively. In addition, this SERS substrate also allows simultaneous detection of the above two antibiotics in one analytical sample with concentrations as low as 10 µg/L for both substances. The proposed method holds promise for addressing concerns related to antibiotic contamination in various environments and food sources, contributing to the advancement of analytical techniques for the trace detection of pharmaceutical residues.

On the Selection of Optimal Balanced Incomplete Block Design using Different Types of Designs

The choice of selecting the best among the types of balanced incomplete block designs (BIBDs) have brought about some controversies in combinatorial design. Prominent methods of constructing BIBD have been proposed by different authors, but none has been juxtaposed the designs with respect to type. In this paper, we proposed the adoption of the concept of optimality criteria such as: A-, D-, E- , and T – optimality criterion to select the best design based on their optimality values using a (15, 15, 7, 7, 3) simulated balanced incomplete block design. A dual design, residue design and derived design were selected as the major three types of BIBDs for investigation. The findings revealed that dual design is the best balanced incomplete block design based on its high optimal value. We discovered that residue and derived design obtained from the simulated balanced incomplete block design by method of block selection and intersection becomes a standard balanced incomplete block design.

Non-Gaussian Liability Distribution for Depression in the General Population.

Unlike depression sum scores, the underlying risk for depression is typically assumed to be normally distributed across the general population. To assess the true empirical shape of depression risk, we created a continuous-valued estimate of the latent depression density, using the Davidian-Curve Item Response Theory (DC-IRT) and the National Health and Nutrition Examination Survey (NHANES) cohorts from 2005 to 2018 (n = 36,244 on the Nine-item Patient Health Questionnaire; PHQ-9). We conducted simulations to investigate the performance of DC-IRT for large samples and realistic items. The method can recover complex latent-risk distributions even when they are not evident from sum scores. However, estimation accuracy for different sample sizes depends on the method of model selection. In addition to full-data analysis, random samples of a few thousand observations were drawn for analysis. The latent shape of depression was left-skewed and bimodal in both investigations, indicating that the latent-normality assumption does not hold for depression.

Optimization of 1H-MRS methods for large-volume acquisition of low-concentration downfield resonances at 3 T and 7 T.

This goal of this study was to optimize spectrally selective 1H-MRS methods for large-volume acquisition of low-concentration metabolites with downfield resonances at 7 T and 3 T, with particular attention paid to detection of nicotinamide adenine dinucleotide (NAD+) and tryptophan. Spectrally selective excitation was used to avoid magnetization-transfer effects with water, and various sinc pulses were compared with a band-selective, uniform response, pure-phase (E-BURP) pulse. Localization using a single-slice selective pulse was compared with voxel-based localization that used three orthogonal refocusing pulses, and low bandwidth refocusing pulses were used to take advantage of the chemical shift displacement of water. A technique for water sideband removal was added, and a method of coil channel combination for large volumes was introduced. Proposed methods were compared qualitatively with previously reported techniques at 7 T. Sinc pulses resulted in reduced water signal excitation and improved spectral quality, with a symmetric, low bandwidth-time product pulse performing best. Single-slice localization allowed shorter TEs with large volumes, enhancing signal, whereas low-bandwidth slice-selective localization greatly reduced the observed water signal. Gradient cycling helped remove water sidebands, and frequency aligning and pruning individual channels narrowed spectral linewidths. High-quality brain spectra of NAD+ and tryptophan are shown in 4 subjects at 3 T. Improved spectral quality with higher downfield signal, shorter TE, lower nuisance signal, reduced artifacts, and narrower peaks was realized at 7 T. These methodological improvements allowed for previously unachievable detection of NAD+ and tryptophan in human brain at 3 T in under 5 min.

Investigation of green synchronous spectrofluorimetric approach for facile sensitive estimation of two co-administered anti-cancer drugs; curcumin and doxorubicin in their laboratory-prepared mixtures, human plasma, and urine

Recently, phytochemicals play an important role in cancer management. Curcumin (CUR), a natural phytochemical, has been co-administered with widespread chemotherapeutic agents such as doxorubicin (DOX) due to its excellent antitumor activity and the ability to lower the adverse reactions and drug resistance cells associated with DOX use. The present study aims to determine DOX and CUR utilizing a label-free, selective, sensitive, and precise synchronous spectrofluorimetric method. The obvious overlap between the emission spectra of DOX and CUR prevents simultaneous estimation of both analytes by conventional spectrofluorimetry. To solve such a problem, synchronous spectrofluorimetric measurements were recorded at Δλ = 20 nm, utilizing ethanol as a diluting solvent. Curcumin was recorded at 442.5 nm, whereas DOX was estimated at 571.5 nm, each at the zero-crossing point of the other one. The developed method exhibited linearity over a concentration range of 0.04–0.40 μg/mL for CUR and 0.05–0.50 μg/mL for DOX, respectively. The values of limit of detection (LOD) were 0.009 and 0.012 µg/mL, while the values of limit of quantitation (LOQ) were 0.028 and 0.037 µg/mL for CUR and DOX, respectively. The adopted approach was carefully validated according to the guidelines of ICH Q2R1. The method was utilized to estimate CUR and DOX in laboratory-prepared mixtures and human biological matrices. It showed a high percentage of recoveries with minimal RSD values. Additionally, three different tools were utilized to evaluate the greenness of the proposed approach.

Machine-learning-based impairment-aware dynamic RMSCA in multi-core elastic optical networks

This paper presents a routing, modulation, spectrum, and core assignment (RMSCA) algorithm for space-division-multiplexing-based elastic optical networks (SDM-EONs) comprising multi-core links. A network state-dependent route and core selection method is proposed using a deep neural network (DNN) classifier. The DNN is trained using a metaheuristic optimization algorithm to predict lightpath suitability, considering the quality of transmission and resource availability. Physical layer impairments, including inter-core crosstalk, amplified spontaneous emission, and Kerr fiber nonlinearities, are considered, and a random forest (RF)-based link noise estimator is proposed. A feature importance selection analysis is provided for all the features considered for the DNN classifier and the RF link noise estimator. The proposed machine-learning-enabled RMSCA approach is evaluated on three network topologies, USNET, NSFNET, and COST-239 with 7-core and 12-core fiber links. It is shown to be superior in terms of blocking probability, bandwidth blocking probability, and acceptable computational speed compared to the standard and published benchmarks at different traffic loads.

A high Ki-67 labeling index and high thyroglobulin doubling rate are significant predictors of excision-site recurrence of papillary thyroid carcinoma following airway resection for locally curative surgery.

Papillary thyroid carcinoma (PTC) occasionally invades the trachea and requires airway resection. Tracheal excision site recurrence (ESR) is a serious problem. We investigated predictors of ESR in patients with PTC who underwent airway resection for locally curative surgery. We enrolled 149 patients with PTC who underwent airway resection (median age at the initial surgery: 67years), including partial-thickness resection (n=73) or full-thickness resection (n=76), for grossly curative surgery. The median postoperative follow-up period was 93months. To date, 11 patients (6.7%) underwent ESR: 6 underwent full-thickness resection and 5 underwent partial-thickness resection. The time to ESR ranged from 14 to 113months (median: 57months) after the initial surgery. None of the 11 ESR patients underwent adjuvant external beam radiotherapy (EBRT) and none of the 4 airway resection patients who underwent EBRT developed ESR. The 5- and 10-year ESR rates were 4.3% and 11.3%, respectively. In the multivariate analysis (forward-backward stepwise selection method), a Ki-67 labeling index (LI) ≥5% (p=0.048) and the thyroglobulin doubling rate (Tg-DR) >0.33/year (p=0.009) (for Tg-antibody negative cases) were independent predictors of ESR. Nine of the 11 patients underwent ESR resection and only one developed a second recurrence. A high Ki-67 LI was a static predictor, and high Tg-DR was a dynamic predictor, of ESR in patients with PTC following airway resection. In such patients, careful postoperative monitoring for ESR is necessary and adjuvant therapies, such as EBRT, may be considered.

Comparative performance of electronegative membrane filtration and automated concentrating pipette for detection of antibiotic resistance genes and microbial markers in river water samples

The target viral and bacterial concentrations in river water are essential for environmental monitoring and public health studies. Filtration-based methods are commonly employed, yet challenges arise due to recoverability and filter pore size. This study aimed to compare the performance of electronegative membrane filtration (EMF) and automated Concentrating Pipette (CP) Select (InnovaPrep) methods for quantifying antibiotic resistance genes (ARGs), mobile genetic elements (MGEs), and bacterial and viral markers in river water samples. Fifty-four river water samples were collected from upstream and downstream locations in a river in Japan. The CP Select method was modified by adding MgCl2 and using different tips. The recovery efficiencies for total coliforms and Escherichia coli were assessed, and class 1 integron-integrase gene (intI1), 16S rRNA, gene encoding sulfonamide resistance (sul1), cross-assembly phage (crAssphage), pepper mild mottle virus (PMMoV), and Escherichia coli gene (sfmD) were detected. CP Select showed recovery efficiencies of 45 %–63 % for total coliforms and 17 %–35 % for E. coli. The intI1, 16S rRNA, sul1, crAssphage, PMMoV, and sfmD concentrations using the modified CP Select method were 10.1 ± 0.5, 8.7 ± 0.2, 7.7 ± 0.2, 6.7 ± 0.2, 5.4 ± 0.2, and 3.5 ± 0.5 log10 copies/L, respectively. Higher intI1 and sul1 concentrations were observed downstream, with the highest contribution percentage (22 % and 21 %) using CP Select or EMF. The modified CP Select method with 0.05 μm tips yielded more quantifiable results for all target genes and greater PMMoV concentrations (p < 0.05). Positive correlations were found among bacterial, ARG/MGE, and viral markers (Spearman's ρ = 0.71 for 16S rRNA and sfmD, 0.88 for intI1 and sul1, and 0.64 for PMMoV and crAssphage). The modified CP Select method demonstrated effective recovery of bacteria and quantification of ARGs, MGEs, and microbial markers in river water. Further studies are required to validate these methods and confirm their applicability in diverse environmental contexts.

Variable Selection in Semi-Functional Partially Linear Regression Models with Time Series Data

This article investigates a variable selection method in semi-functional partially linear regression (SFPLR) models for strong α-mixing functional time series data. We construct penalized least squares estimators for unknown parameters and unknown link functions in our models. Under some regularity assumptions, we establish the asymptotic convergence rate and asymptotic distribution for the proposed estimators. Furthermore, we make a comparison of our variable selection method with the oracle method without variable selection in simulation studies and an electricity consumption data analysis. Simulation experiments and real data analysis results indicate that the variable selection method performs well at extracting the primary information and reducing dimensionality.

Supervised star, galaxy, and QSO classification with sharpened dimensionality reduction

We explored the use of broadband colors to classify stars, galaxies, and quasi-stellar objects (QSOs). Specifically, we applied sharpened dimensionality reduction (SDR)-aided classification to this problem, with the aim of enhancing cluster separation in the projections of high-dimensional data clusters to allow for better classification performance and more informative projections. The main objective of this work was to apply SDR to large sets of broadband colors derived from the CPz catalog to obtain projections with clusters of star, galaxy, and QSO data that exhibit a high degree of separation. The SDR method achieves this by combining density-based clustering with conventional dimensionality-reduction techniques. To make SDR scalable and have the ability to project samples using the earlier-computed projection, we used a deep neural network trained to reproduce the SDR projections. Subsequently classification was done by applying a $k$-nearest neighbors ($k$-NN) classifier to the sharpened projections. Based on a qualitative and quantitative analysis of the embeddings produced by SDR, we find that SDR consistently produces accurate projections with a high degree of cluster separation. A number of projection performance metrics are used to evaluate this separation, including the trustworthiness, continuity, Shepard goodness, and distribution consistency metrics. Using the $k$-NN classifier and consolidating the results of various data sets, we obtain precisions of 99.7&lt;!PCT!&gt;, 98.9&lt;!PCT!&gt;, and 98.5&lt;!PCT!&gt; for classifying stars, galaxies, and QSOs, respectively. Furthermore, we achieve completenesses of 97.8&lt;!PCT!&gt;, 99.3&lt;!PCT!&gt;, and 86.8&lt;!PCT!&gt;, respectively. In addition to classification, we explore the structure of the embeddings produced by SDR by cross-matching with data from Gaia DR3, Galaxy Zoo 1, and a catalog of specific star formation rates, stellar masses, and dust luminosities. We discover that the embeddings reveal astrophysical information, which allows one to understand the structure of the high-dimensional broadband color data in greater detail. We find that SDR-aided star, galaxy, and QSO classification performs comparably to another unsupervised learning method using hierarchical density-based spatial clustering of applications with noise (HDBSCAN) but offers advantages in terms of scalability and interpretability. Furthermore, it outperforms traditional color selection methods in terms of QSO classification performance. Overall, we demonstrate the potential of SDR-aided classification to provide an accurate and physically insightful classification of astronomical objects based on their broadband colors.