Shubh Saraswat

Mr. Shubh Saraswat

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
7 May 2024 in Nutrients
Verified
Journal Article
Open access
PDF file
GKGabriella H. Kalantar
Gabriella H. Kalantar
SSShubh Saraswat
Shubh Saraswat
SSSara SantaCruz-Calvo
Sara SantaCruz-Calvo
FGFatemeh Gholamrezaeinejad
Fatemeh Gholamrezaeinejad
AJAida Javidan
Aida Javidan
BNBarbara S Nikolajczyk
Barbara S Nikolajczyk
Cytokines produced by peripheral T-helper 1/17 cells disproportionately contribute to the inflammation (i.e., metaflammation) that fuels type 2 diabetes (T2D) pathogenesis. Shifts in the nutrient milieu could influence inflammation through changes in T-cell metabolism. We aimed to determine whether changes in glucose utilization alter cytokine profiles in T2D. Peripheral blood mononuclear cells (PBMCs), CD4+ T-cells, and CD4+CD25− T-effector (Teff) cells were isolated from age-matched humans classified by glycemic control and BMI. Cytokines secreted by CD3/CD28-stimulated PBMCs and Teff were measured in supernatants with multiplex cytokine assays and a FLEXMAP-3D. Metabolic activity of stimulated CD4+ T-cells was measured by a Seahorse XFe96 analyzer. In this study, we demonstrated that T-cell stimulated PBMCs from non-fasted people with T2D produced higher amounts of cytokines compared to fasting. Although dysglycemia characterizes T2D, cytokine production by PBMCs or CD4+ T-cells in T2D was unaltered by hyperglycemic media. Moreover, pharmacological suppression of mitochondrial glucose oxidation did not change T-cell metabolism in T2D, yet enhanced cytokine competency. In conclusion, fasting and glucose metabolism differentially impact peripheral inflammation in human T2D, suggesting that glucose, along with fatty acid metabolites per our previous work, partner to regulate metaflammation. These data expose a major disconnect in the use of glycemic control drugs to target T2D-associated metaflammation.

1 March 2026 in Computational and Structural Biotechnology Journal
Verified
Journal Article
Open access
SSShubh Saraswat
Shubh Saraswat
BNBira Arumndari Nurrahma
Bira Arumndari Nurrahma
PKPhilip A Kern
Philip A Kern
BNBarbara S Nikolajczyk
Barbara S Nikolajczyk
XZXiaohua Douglas Zhang
Xiaohua Douglas Zhang
Background: Multiplex cytokine assays are widely used for biomarker discovery and disease profiling, but most existing web tools are optimized for metabolomics and require substantial data reformatting. A reproducible, end-to-end computational platform tailored to cytokine analysis is needed. Methods: We developed CytokineProfile, a web-based analytic tool that accepts standard tabular formats and guides users through data preview, quality control, missing-value handling, and filtering. Analytical modules include exploratory visualization (boxplots, correlation matrices, volcano plots, and dual-flashlight plots combining log fold change with strictly standardized mean difference), univariate tests (analysis of variance and normality-informed 2-sample tests), multivariate modeling (principal component analysis, sparse partial least squares discriminant analysis, and multilevel/multivariate integration variants), and machine-learning classifiers (extreme gradient boosting and random forest) with feature-importance outputs. Results: In a case study using anti-CD3/anti-CD28-stimulated peripheral blood mononuclear cell cytokine profiles from subjects with prediabetes at 20 and 72 h (33 samples per time point), CytokineProfile supported same-dataset comparison with MetaboAnalyst and CytokineExplore, generated consistent group separation in supervised analyses, and provided complementary volcano, dual-flashlight, and machine-learning-based feature results. Benchmarking on datasets up to 1,000 samples and 100 cytokines showed practical runtime and memory usage for representative workflows. Conclusion: CytokineProfile provides a reproducible, user-friendly platform integrating quality control, statistical analysis, visualization, and machine learning. The tool streamlines cytokine biomarker discovery and interpretation, supporting rigorous analysis of complex datasets.

3 April 2025
Verified
Dataset
Closed access
SSShubh Saraswat
Shubh Saraswat
XZXiaohua Douglas Zhang
Xiaohua Douglas Zhang
Provides comprehensive cytokine profiling analysis through quality control using biologically meaningful cutoffs on raw cytokine measurements and by testing for distributional symmetry to recommend appropriate transformations. Offers exploratory data analysis with summary statistics, enhanced boxplots, and barplots, along with univariate and multivariate analytical capabilities for in-depth cytokine profiling such as Principal Component Analysis based on Andrzej Maćkiewicz and Waldemar Ratajczak (1993) , Sparse Partial Least Squares Discriminant Analysis based on Lê Cao K-A, Boitard S, and Besse P (2011) , Random Forest based on Breiman, L. (2001) , and Extreme Gradient Boosting based on Tianqi Chen and Carlos Guestrin (2016) .

3 February 2026
Verified
Dataset
Closed access
SSShubh Saraswat
Shubh Saraswat
HSHasin Shahed Shad
Hasin Shahed Shad
XZXiaohua Douglas Zhang
Xiaohua Douglas Zhang
Evaluates predictive performance under feature-level missingness in repeated-measures continuous glucose monitoring-like data. The benchmark injects missing values at user-specified rates, imputes incomplete feature matrices using an iterative chained-equations approach inspired by multivariate imputation by chained equations (MICE; Azur et al. (2011) ), fits Random Forest regression models (Breiman (2001) ) and k-nearest-neighbor regression models (Zhang (2016) ), and reports mean absolute percentage error and R-squared across missingness rates.

5 February 2026 in Ecotoxicology and Environmental Safety
Verified
Journal Article
Open access
KSKrystal M. Sewell
Krystal M. Sewell
KOKelly Oriakhi
Kelly Oriakhi
STSara Y. Ngo Tenlep
Sara Y. Ngo Tenlep
MWMegan Weaver
Megan Weaver
MBMansura Shahad Bawa
Mansura Shahad Bawa
CRCetewayo S. Rashid
Cetewayo S. Rashid
Organophosphate flame retardants (OPFRs) are used in many consumer products, and their presence in our environment raises concerns about potential health dangers. In humans, exposure to OPFRs, such as tris(1,3-dichloro-2-propyl) phosphate (TDCPP) is associated with metabolic disorders such as obesity, hyperglycemia, and insulin resistance. Our current study seeks to determine how TDCPP exposure affects body composition, glucose homeostasis, and energy balance in a rodent model. C57Bl/6 J mice were divided into two weight-matched groups and fed purified low phytoestrogen diets containing either 0.1 % DMSO (control) or TDCPP 415 mg/kg resulting in an approximate exposure of 50 mg/kg/day. Both male and female mice were used. Promethion metabolic cages were used to measure the rates of oxygen consumption and CO2 generation, food intake, and locomotion. Body composition was measured using nuclear magnetic resonance relaxometry. Mice underwent a glucose tolerance test, and hyperinsulinemic euglycemic clamp was used to assess insulin sensitivity. Male TDCPP-treated mice showed higher food intake and energy expenditure without changes in locomotion or energy balance. TDCPP tended to differentially affect respiratory exchange ratio, which was increased in males and decreased in females. Male mice showed impaired glucose tolerance, whereas female mice did not. Hyperinsulinemic clamp revealed no significant decrease in the glucose infusion rate of TDCPP-exposed animals. However, hepatic glucose production during hyperinsulinemia was increased compared to controls signifying liver-specific insulin resistance. This study supports the cause for concern related to detrimental metabolic health effects of TDCPP exposure.

16 July 2026
Verified
Dataset
Closed access
SSShubh Saraswat
Shubh Saraswat
HSHasin Shahed Shad
Hasin Shahed Shad
XZXiaohua Douglas Zhang
Xiaohua Douglas Zhang
Imputes missing glucose values in repeated-measures continuous glucose monitoring (CGM) data. Workflows create time-series features from raw timestamps, support model selection, and return the user's original columns plus an imputed glucose column. Methods include multiple imputation by chained equations using 'mice' (Azur et al. (2011) ), Random Forest regression using 'ranger' (Breiman (2001) ), k-nearest-neighbor regression using 'FNN' (Zhang (2016) ), 'XGBoost' using 'xgboost' (Chen and Guestrin (2016) ), 'LightGBM' using 'lightgbm' (Ke et al. (2017) ), and ARIMA forecasting using 'forecast' (Hyndman and Khandakar (2008) ). A 'Python'-compatible backend uses 'reticulate' to call 'pandas', 'scikit-learn', 'statsmodels', 'xgboost', and optional 'lightgbm'.