Bradley E. Aouizerat

Abstract

Purpose: Sexual and gender minority (SGM) people are at greater risk for substance use than heterosexual and cisgender people, but most prior work is limited by cross-sectional analyses or the examination of single substance use. This study examined substance use over time among SGM people to identify patterns of polysubstance use at the intersection of sex and gender. Methods: Data were collected annually over 4 years from SGM respondents (n = 11,822) in The Population Research in Identity and Disparities for Equality (PRIDE) Study. Differences in substance use patterns (any prior 30-day use of 15 substances) by gender subgroup were examined with latent class analysis, and multinomial regression models tested relationships between gender subgroup and substance use. Results: Eight classes of substance use were observed. The three most common patterns were low substance use (49%), heavy episodic alcohol use (≥5 alcoholic drinks on one occasion) with some cannabis and tobacco use (14%), and cannabis use with some tobacco and declining heavy episodic alcohol use (13%). Differences observed included lower odds of patterns defined by heavy episodic alcohol use with some cannabis and tobacco use in all gender subgroups relative to cisgender men and persons with low substance use (odds ratios [ORs] 0.26-0.60). Gender expansive people assigned female at birth, gender expansive people assigned male at birth, and transgender men had greater odds of reporting cannabis use with small percentages of heavy episodic alcohol and tobacco use (ORs: 1.41-1.60). Conclusion: This study suggests that there are unique patterns of polysubstance use over time among gender subgroups of SGM people.

Abstract

Co-occurrence of injection drug use (IDU) and hepatitis C virus infection (HCV) is common in people living with HIV (PLWH) and leads to significantly increased mortality. Epigenetic clocks derived from DNA methylation (DNAm) are associated with disease progression and all-cause mortality. In this study, we hypothesized that epigenetic age mediates the relationships between the co-occurrence of IDU and HCV with mortality risk among PLWH. We tested this hypothesis in the Veterans Aging Cohort Study (n = 927) by using four established epigenetic clocks of DNAm age (i.e., Horvath, Hannum, Pheno, Grim). Compared to individuals without IDU and HCV (IDU-HCV-), participants with IDU and HCV (IDU+HCV+) showed a 2.23-fold greater risk of mortality estimated using a Cox proportional hazards model (hazard ratio: 2.23; 95% confidence interval: 1.62–3.09; p = 1.09E–06). IDU+HCV+ was associated with a significantly increased epigenetic age acceleration (EAA) measured by 3 out of 4 epigenetic clocks, adjusting for demographic and clinical variables (Hannum: p = 8.90E–04, Pheno: p = 2.34E–03, Grim: p = 3.33E–11). Furthermore, we found that epigenetic age partially mediated the relationship between IDU+HCV+ and all-cause mortality, up to a 13.67% mediation proportion. Our results suggest that comorbid IDU with HCV increases EAA in PLWH that partially mediates the increased mortality risk.

Abstract

Purpose: To examine if gut microbial taxa abundances and predicted functional pathways correlate with Bristol Stool Form Scale (BSFS) classification at the end of neoadjuvant chemotherapy and radiation therapy (CRT) for rectal cancer. Methods: Rectal cancer patients (n = 39) provided stool samples for 16S rRNA gene sequencing. Stool consistency was evaluated using the BSFS. Gut microbiome data were analyzed using QIIME2. Correlation analysis were performed in R. Results: At the genus level, Staphylococcus positively correlates (Spearman’s rho = 0.26), while Anaerofustis, Roseburia, Peptostreptococcaceae unclassified, Ruminococcaceae UBA1819, Shuttleworthia, Ca. Soleaferrea, Anaerostignum, Oscillibacter, and Akkermansia negatively correlate with BSFS scores (Spearman’s rho −0.20 to −0.42). Predicted pathways, including mycothiol biosynthesis and sucrose degradation III (sucrose invertase), were positively correlated with BSFS (Spearman’s rho = 0.03–0.21). Conclusion: The data support that in rectal cancer patients, stool consistency is an important factor to include in microbiome studies. Loose/liquid stools may be linked to Staphylococcus abundance and to mycothiol biosynthesis and sucrose degradation pathways.

Abstract

Aims/Hypothesis: Our prior analysis of the Diabetes Prevention Program study identified a subset of five miRNAs that predict incident type 2 diabetes. The purpose of this study was to identify mRNAs and biological pathways targeted by these five miRNAs to elucidate potential mechanisms of risk and responses to the tested interventions. Methods: Using experimentally validated data from miRTarBase version 8.0 and R (2021), we identified mRNAs with strong evidence to be regulated by individual or combinations of the five predictor miRNAs. Overrepresentation of the mRNA targets was assessed in pathways from the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotation database. Results: The five miRNAs targeted 167 pathways and 122 mRNAs. Nine of the pathways have known associations with type 2 diabetes: Insulin signaling, Insulin resistance, Diabetic cardiomyopathy, Type 2 diabetes, AGE-RAGE signaling in diabetic complications, HIF-1 signaling, TGF-beta signaling, PI3K/Akt signaling, and Adipocytokine signaling pathways. Vascular endothelial growth factor A (VEGFA) has prior genetic associations with risk for type 2 diabetes and was the most commonly targeted mRNA for this set of miRNAs. Conclusions/Interpretation: These findings show that miRNA predictors of incident type 2 diabetes target mRNAs and pathways known to underlie risk for type 2 diabetes. Future studies should evaluate miRNAs as potential therapeutic targets for preventing and treating type 2 diabetes.

Abstract

Context: MicroRNAs (miRs) are short (ie, 18-26 nucleotide) regulatory elements of messenger RNA translation to amino acids. Objective: The purpose of this study was to assess whether miRs are predictive of incident type 2 diabetes (T2D) in the Diabetes Prevention Program (DPP) trial. Methods: This was a secondary analysis (n = 1000) of a subset of the DPP cohort that leveraged banked biospecimens to measure miRs. We used random survival forest and Lasso methods to identify the optimal miR predictors and the Cox proportional hazards to model time to T2D overall and within intervention arms. Results: We identified 5 miRs (miR-144, miR-186, miR-203a, miR-205, miR-206) that constituted the optimal predictors of incident T2D after adjustment for covariates (hazard ratio [HR] 2.81, 95% CI 2.05, 3.87; P < .001). Predictive risk scores following cross-validation showed the HR for the highest quartile risk group compared with the lowest quartile risk group was 5.91 (95% CI 2.02, 17.3; P < .001). There was significant interaction between the intensive lifestyle (HR 3.60, 95% CI 2.50, 5.18; P < .001) and the metformin (HR 2.72; 95% CI 1.47, 5.00; P = .001) groups compared with placebo. Of the 5 miRs identified, 1 targets a gene with prior known associations with risk for T2D. Conclusion: We identified 5 miRs that are optimal predictors of incident T2D in the DPP cohort. Future directions include validation of this finding in an independent sample in order to determine whether this risk score may have potential clinical utility for risk stratification of individuals with prediabetes, and functional analysis of the potential genes and pathways targeted by the miRs that were included in the risk score.

Abstract

Introduction: Integrated transcriptome and microRNA differential gene expression (DEG) analyses may help to explain type 2 diabetes (T2D) pathogenesis in at-risk populations. The purpose of this study was to characterize DEG in banked biospecimens from underactive adult participants who responded to a randomized clinical trial measuring the effects of lifestyle interventions on T2D risk factors. DEGs were further examined within the context of annotated biological pathways. Methods: Participants (n = 52) in a previously completed clinical trial that assessed a 12-week behavioural intervention for T2D risk reduction were included. Participants who showed >6mg/dL decrease in fasting blood glucose were identified as responders. Gene expression was measured by RNASeq, and overrepresentation analysis within KEGG pathways and weighted gene correlation network analysis (WGCNA) were performed. Results: No genes remained significantly differentially expressed after correction for multiple comparisons. One module derived by WGCNA related to body mass index was identified, which contained genes located in KEGG pathways related to known mechanisms underlying risk for T2D as well as pathways related to neurodegeneration and protein misfolding. A network analysis showed indirect connections between genes in this module and islet amyloid polypeptide (IAPP), which has previously been hypothesized as a mechanism for T2D. Discussion: We validated prior studies that showed pathways related to metabolism, inflammation/immunity, and endocrine/hormone function are related to risk for T2D. We identified evidence for new potential mechanisms that include protein misfolding. Additional studies are needed to determine whether these are potential therapeutic targets to decrease risk for T2D.

Abstract

Background: Accurate prediction of risk for chronic diseases like type 2 diabetes (T2D) is challenging due to the complex underlying etiology. Integration of more complex data types from sensors and leveraging technologies for collection of -omics datasets may provide greater insights into the specific risk profile for complex diseases. Methods: We performed a literature review to identify feature selection methods and machine learning models for prediction of weight loss in a previously completed clinical trial (NCT02278939) of a behavioral intervention for weight loss in Filipinos at risk for T2D. Features included demographic and clinical characteristics, dietary factors, physical activity, and transcriptomics. Results: We identified four feature selection methods: Correlation-based Feature Subset Selection (CfsSubsetEval) with BestFirst, Kolmogorov–Smirnov (KS) test with correlation featureselection (CFS), DESeq2, and max-relevance-min-relevance (MRMR) with linear forward search and mutual information (MI) and four machine learning algorithms: support vector machine, decision tree, random forest, and extra trees that are applicable to prediction of weight loss using the specified feature types. Conclusion: More accurate prediction of risk for T2D and other complex conditions may be possible by leveraging complex data types from sensors and -omics datasets. Emerging methods for feature selection and machine learning algorithms make this type of modeling feasible.

Abstract

Background: Type 2 diabetes (T2D) has an immense disease burden, affecting millions of people worldwide and costing billions of dollars in treatment. As T2D is a multifactorial disease with both genetic and nongenetic influences, accurate risk assessments for patients are difficult to perform. Machine learning has served as a useful tool in T2D risk prediction, as it can analyze and detect patterns in large and complex data sets like that of RNA sequencing. However, before machine learning can be implemented, feature selection is a necessary step to reduce the dimensionality in high-dimensional data and optimize modeling results. Different combinations of feature selection methods and machine learning models have been used in studies reporting disease predictions and classifications with high accuracy. Objective: The purpose of this study was to assess the use of feature selection and classification approaches that integrate different data types to predict weight loss for the prevention of T2D. Methods: The data of 56 participants (ie, demographic and clinical factors, dietary scores, step counts, and transcriptomics) were obtained from a previously completed randomized clinical trial adaptation of the Diabetes Prevention Program study. Feature selection methods were used to select for subsets of transcripts to be used in the selected classification approaches: support vector machine, logistic regression, decision trees, random forest, and extremely randomized decision trees (extra-trees). Data types were included in different classification approaches in an additive manner to assess model performance for the prediction of weight loss. Results: Average waist and hip circumference were found to be different between those who exhibited weight loss and those who did not exhibit weight loss (P=.02 and P=.04, respectively). The incorporation of dietary and step count data did not improve modeling performance compared to classifiers that included only demographic and clinical data. Optimal subsets of transcripts identified through feature selection yielded higher prediction accuracy than when all available transcripts were included. After comparison of different feature selection methods and classifiers, DESeq2 as a feature selection method and an extra-trees classifier with and without ensemble learning provided the most optimal results, as defined by differences in training and testing accuracy, cross-validated area under the curve, and other factors. We identified 5 genes in two or more of the feature selection subsets (ie, CDP-diacylglycerol-inositol 3-phosphatidyltransferase [CDIPT], mannose receptor C type 2 [MRC2], PAT1 homolog 2 [PATL2], regulatory factor X-associated ankyrin containing protein [RFXANK], and small ubiquitin like modifier 3 [SUMO3]). Conclusions: Our results suggest that the inclusion of transcriptomic data in classification approaches for prediction has the potential to improve weight loss prediction models. Identification of which individuals are likely to respond to interventions for weight loss may help to prevent incident T2D. Out of the 5 genes identified as optimal predictors, 3 (ie, CDIPT, MRC2, and SUMO3) have been previously shown to be associated with T2D or obesity.

Abstract

MicroRNAs (miRs) may contribute to disease etiology by influencing gene expression. Numerous databases are available for miR target prediction and validation, but their functionality is varied, and outputs are not standardized. The purpose of this review is to identify and describe databases for cataloging validated miR targets. Using Tools4miRs and PubMed, we identified databases with experimentally validated targets, human data, and a focus on miR-messenger RNA (mRNA) interactions. Data were extracted about the number of times each database was cited, the number of miRs, the target genes, the interactions per database, experimental methodology and key features of each database. The search yielded 10 databases, which in order of most cited to least were: miRTarBase, starBase/The Encyclopedia of RNA Interactomes, DIANA-TarBase, miRWalk, miRecords, miRGator, miRSystem, miRGate, miRSel and targetHub. Findings from this review suggest that the information presented within miR target validation databases can be enhanced by adding features such as flexibility in performing queries in multiple ways, downloadable data, ongoing updates and integrating tools for further miR-mRNA target interaction analysis. This review is designed to aid researchers, especially those new to miR bioinformatics tools, in database selection and to offer considerations for future development and upkeep of validation tools. Database URL http://mirtarbase.cuhk.edu.cn/