Gary Krzyzanowski

Extracellular vesicle (EV) transcriptomic data provides a high-dimensional representation of cellular states but remains challenging to interpret due to noise, redundancy, and limited sample sizes. Most existing approaches rely on supervised differential expression analyses, which can be biased and may fail to capture latent structure in small datasets. In this study, we propose an unsupervised machine learning framework based on non-negative matrix factorization (NMF) to identify latent expression programs from EV mRNA profiles. A structured preprocessing pipeline combining expression filtering, variance selection, ANOVA-based feature selection, and correlation pruning was used to reduce dimensionality and improve signal quality prior to matrix factorization. NMF was applied to decompose the data into interpretable gene modules and sample-specific activation patterns. Model selection was performed using reconstruction error and component stability across multiple initializations. Candidate features were prioritized using a composite ranking score integrating module loadings, group-level expression differences, and model stability. The approach identified a stable low-rank representation capturing dominant patterns in the data and a compact set of informative features. These results demonstrate that unsupervised matrix factorization can effectively extract structured, interpretable signals from small-scale transcriptomic datasets and provide a robust framework for feature prioritization and representation learning in high-dimensional biological data.

A distinct form of chronic kidney disease of unknown etiology (CKDu) has emerged in tropical regions of Sri Lanka, predominantly affecting individuals aged 30–60 years in the North Central Province. Unlike conventional chronic kidney disease (CKD), CKDu occurs independently of diabetes or hypertension and is characterized by tubulointerstitial damage, including tubular atrophy, interstitial inflammation, and fibrosis. Epidemiological studies showed familial clustering, suggesting an underlying genetic predisposition. This study aimed to identify genetic variants associated with CKDu in Sri Lankan populations using whole-exome sequencing (WES). Eighty-six individuals (47 CKDu patients and 39 controls) were recruited from endemic and non-endemic regions. Physiological, biochemical, and geographic parameters were recorded. DNA extracted from blood was subjected to WES to identify variants associated with CKDu. Results: A total of 171 unique variants across 121 genes were identified. Among the most prevalent genes were ATXN3, LFNG, PNLDC1, LINC02456, and HLA-DRB1. In the case–control comparison, only LFNG showed statistically significant enrichment in affected individuals, whereas signals in ATXN3, PNLDC1, and LINC02456 were not statistically significant, but have an association with renal dysfunction, and thus are included as hypothesis-generating variant observations. HLA-DRB1 variants showed trends toward a protective haplotype. LFNG showed the greatest prevalence in affected individuals (71.7%), followed by PNLDC1 (63%), ATXN3 (56%), FIP1L1 (41%), and HLA-DRB1 (32%). Conclusion: Findings suggest genetic variants in combination with environmental factors may contribute to CKDu susceptibility in the Sri Lankan population. We underscore the multi-factorial nature of CKDu and highlight the need for integrative genomic and environmental research to elucidate disease mechanisms and inform targeted prevention strategies.

Background/Objectives: Ovarian cancer is a heterogeneous malignancy with molecular subtypes that strongly influence prognosis and therapy. High-dimensional mRNA data can capture this biological diversity, but its complexity and noise limit robust subtype characterization. Furthermore, current classification approaches often fail to reflect subtype-specific transcriptional programs, underscoring the need for computational strategies that reduce dimensionality and identify discriminative molecular features. Methods: We designed a multi-stage feature selection and network analysis framework tailored for high-dimensional transcriptomic data. Starting with ~65,000 mRNA features, we applied unsupervised variance-based filtering and correlation pruning to eliminate low-information genes and reduce redundancy. The applied supervised Select-K Best filtering further refined the feature space. To enhance robustness, we implemented a hybrid selection strategy combining recursive feature elimination (RFE) with random forests and LASSO regression to identify discriminative mRNA features. Finally, these features were then used to construct a gene co-expression similarity network. Results: This pipeline reduced approximately 65,000 gene features to a subset of 83 discriminative transcripts, which were then used for network construction to reveal subtype-specific biology. The analysis identified four distinct groups. One group exhibited classical high-grade serous features defined by TP53 mutations and homologous recombination deficiency, while another was enriched for PI3K/AKT and ARID1A-associated signaling consistent with clear cell and endometrioid-like biology. A third group displayed drug resistance-associated transcriptional programs with receptor tyrosine kinase activation, and the fourth demonstrated a hybrid profile bridging serous and endometrioid expression modules. Conclusions: This pilot study shows that combining unsupervised and supervised feature selection with network modeling enables robust stratification of ovarian cancer subtypes.

Usher syndrome, a rare genetic disorder causing both hearing and vision loss, presents significant diagnostic and therapeutic challenges due to its complex genetic basis. The identification of reliable biomarkers for early detection and intervention is crucial for improving patient outcomes. In this study, we present a machine learning-based hybrid sequential feature selection approach to identify key mRNA biomarkers associated with Usher syndrome. Beginning with a dataset of 42,334 mRNA features, our approach successfully reduced dimensionality and identified 58 top mRNA biomarkers that distinguish Usher syndrome from control samples. We employed a combination of feature selection techniques, including variance thresholding, recursive feature elimination, and Lasso regression, integrated within a nested cross-validation framework. The selected biomarkers were further validated using multiple machine learning models, including Logistic Regression, Random Forest, and Support Vector Machines, demonstrating robust classification performance. To assess the biological relevance of the computationally identified mRNA biomarkers, we experimentally validated candidates from the top 10 selected mRNAs using droplet digital PCR (ddPCR). The ddPCR results were consistent with expression patterns observed in the integrated transcriptomic metadata, reinforcing the credibility of our machine learning-driven biomarker discovery framework. Our findings highlight the potential of machine learning-driven biomarker discovery to enhance the detection of Usher syndrome.

Usher syndrome (USH) is a rare genetic disorder affecting vision, hearing, and balance. Identifying reliable biomarkers is crucial for early diagnosis and understanding disease mechanisms. MicroRNAs (miRNAs), key regulators of gene expression, hold promise as biomarkers for USH. This study aimed to identify a minimal subset of miRNAs that could serve as biomarkers to effectively differentiate USH from controls. We employed ensemble feature selection techniques to select the top miRNAs appearing in at least three algorithms. Machine learning models were trained and tested using this subset, followed by validation on an independent 10% sample. Our approach identified 10 key miRNAs as potential biomarkers for USH. To further validate their biological relevance, we conducted pathway analysis, which revealed significant pathways associated with USH. Furthermore, our approach achieved high classification performance, with an accuracy of 97.7%, sensitivity of 98%, specificity of 92.5%, F1 score of 95.8%, and an AUC of 97.5%. These findings demonstrate that combining ensemble feature selection with machine learning provides a robust strategy for miRNA biomarker discovery, advancing USH diagnosis and molecular understanding.

Background: The thioredoxin system (TrxS) is crucial for maintaining redox balance by regulating cellular thiol levels and is involved in various biological processes, including cancer progression. Thioredoxin reductase (TrxR), a key component of TrxS, reduces oxidized thioredoxin (Trx) using NADPH. This study investigates the inhibitory effects of 2-bromo-2-nitro-1,3-propanediol (Bronopol, BP), a preservative, on TrxR activity and its impact on cellular thiols and cell viability. Methods: Purified recombinant TrxR and noncancer and cancer cells were treated with different concentrations of BP and TrxR activity measured. BP-treated cells were examined for effects of BP on total cellular thiol level and GSH/GSSG ratio. Results: BP effectively inhibited TrxR in a dose-dependent manner, an effect that was reversible with dithiothreitol (DTT). BP treatment significantly reduced total thiol levels, decreased the GSH/GSSG ratio, and increased reactive oxygen species (ROS) in cells. Additionally, BP decreased cell viability and induced apoptosis, as indicated by morphological changes and increased c-fos mRNA expression. Conclusions: These findings highlight BP’s potential as a TrxR inhibitor and its cytotoxicity toward both noncancer and cancer cells. The observed effects—TrxR inhibition, thiol oxidation, GSH/GSSG imbalance, and ROS accumulation—may underlie BP’s cytotoxicity. Further research is needed to explore the precise molecular mechanisms by which BP exerts these effects.

Background: Usher syndrome (USH) is a rare genetic disorder that affects both hearing and vision. It presents in three clinical types—USH1, USH2, and USH3—with varying onset, severity, and disease progression. Existing diagnostics primarily rely on genetic profiling to identify variants in USH genes; however, accurate detection before symptom onset remains a challenge. MicroRNAs (miRNAs), which regulate gene expression, have been identified as potential biomarkers for disease. The aim of this study is to develop a data-driven system for the identification of USH using miRNA expression profiles. Methods: We collected microarray miRNA-expression data from 17 samples, representing four patient-derived USH cell lines and a non-USH control. Supervised feature selection was utilized to identify key miRNAs that differentiate USH cell lines from the non-USH control. Subsequently, a network model was constructed by measuring pairwise correlations based on these identified features. Results: The proposed system effectively distinguished between control and USH samples, demonstrating high accuracy. Additionally, the model could differentiate between the three USH types, reflecting its potential and sensitivity beyond the primary identification of affected subjects. Conclusions: This approach can be used to detect USH and differentiate between USH subtypes, suggesting its potential as a future base model for the identification of Usher syndrome.

Usher syndrome (USH) is an inherited disorder characterized by sensorineural hearing loss (SNHL), retinitis pigmentosa (RP)-related vision loss, and vestibular dysfunction. USH presents itself as three distinct clinical types, 1, 2, and 3, with no biomarker for early detection. This study aimed to explore whether microRNA (miRNA) expression in USH cell lines is dysregulated compared to the miRNA expression pattern in a cell line derived from a healthy human subject. Lymphocytes from USH patients and healthy individuals were isolated and transformed into stable cell lines using Epstein–Barr virus (EBV). DNA from these cell lines was sequenced using a targeted panel to identify gene variants associated with USH types 1, 2, and 3. Microarray analysis was performed on RNA from both USH and control cell lines using NanoString miRNA microarray technology. Dysregulated miRNAs identified by the microarray were validated using droplet digital PCR technology. DNA sequencing revealed that two USH patients had USH type 1 with gene variants in USH1B (MYO7A) and USH1D (CDH23), while the other two patients were classified as USH type 2 (USH2A) and USH type 3 (CLRN-1), respectively. The NanoString miRNA microarray detected 92 differentially expressed miRNAs in USH cell lines compared to controls. Significantly altered miRNAs exhibited at least a twofold increase or decrease with a p value below 0.05. Among these miRNAs, 20 were specific to USH1, 14 to USH2, and 5 to USH3. Three miRNAs that are known as miRNA-183 family which are crucial for inner ear and retina development, have been significantly downregulated as compared to control cells. Subsequently, droplet digital PCR assays confirmed the dysregulation of the 12 most prominent miRNAs in USH cell lines. This study identifies several miRNA signatures in USH cell lines which may have potential utility in Usher syndrome identification.

Circulating miRNA-451a has recently emerged as an important biomolecule with potential clinical values as a diagnostic marker for several diseases. However, to be used as such, it is critical to accurately assay miRNA-451a in the clinic. Yet, pre-analytical factors that can affect such an error-free clinical assay of miRNA-451a have not been explored. This study aimed at investigating several of these pre-analytical factors that may affect the accurate quantification of miRNA-451a in human blood samples. We initially evaluated miRNA-451a levels in red blood cells (RBCs), white blood cells (WBCs), platelets, and plasma by droplet digital PCR (ddPCR). Next, we quantified miRNA-451a levels in whole blood or plasma stored at different temperatures for different time periods by ddPCR and used hemolyzed plasma samples to study the effects of hemolysis on plasma miRNA-451a concentration. We found that WBCs and platelets contain trace amounts of miRNA-451a and that about 99.9% of miRNA-451a in the blood are localized in RBCs. Hemolyzed plasma showed 58.5-fold increase in miRNA-451 concentration as compared to non-hemolyzed plasma. Importantly, plasma stored at room temperature (RT) or 4ºC showed a steady statistically significant decrease in miRNA-451a levels over a 7-day period whereas plasma samples stored at -20 or -80ºC for up to 7 days showed stable miRNA-451a levels. Together, our data indicate that hemolysis and blood storage at RT or 4ºC may have significant negative effects on the levels of circulating miRNA-451a.