Estimation of Sobol's Sensitivity Indices under Generalized Linear Models

Targeted preventive vaccination campaigns to reduce Ebola outbreaks: An individual-based modeling study

INTRODUCTION

Nonpharmaceutical interventions (NPI) and ring vaccination (i.e., vaccination that primarily targets contacts and contacts of contacts of Ebola cases) are currently used to reduce the spread of Ebola during outbreaks. Because these measures are typically initiated after an outbreak is declared, they are limited by real-time implementation challenges. Preventive vaccination may provide a complementary option to help protect communities against unpredictable outbreaks. This study aimed to assess the impact of preventive vaccination strategies when implemented in conjunction with NPI and ring vaccination.

METHODS

A spatial-explicit, individual-based model (IBM) that accounts for heterogeneity of human contact, human movement, and timing of interventions was built to represent Ebola transmission in the Democratic Republic of the Congo. Simulated preventive vaccination strategies targeted healthcare workers (HCW), frontline workers (FW), and the general population (GP) with varying levels of coverage (lower coverage: 30% of HCW/FW, 5% of GP; higher coverage: 60% of HCW/FW, 10% of GP) and efficacy (lower efficacy: 60%; higher efficacy: 90%).

RESULTS

The IBM estimated that the addition of preventive vaccination for HCW reduced cases, hospitalizations, and deaths by ∼11 % to ∼25 % compared with NPI + ring vaccination alone. Including HCW and FW in the preventive vaccination campaign yielded ∼14 % to ∼38 % improvements in epidemic outcomes. Further including the GP yielded the greatest improvements, with ∼21 % to ∼52 % reductions in epidemic outcomes compared with NPI + ring vaccination alone. In a scenario without ring vaccination, preventive vaccination reduced cases, hospitalizations, and deaths by ∼28 % to ∼59 % compared with NPI alone. In all scenarios, preventive vaccination reduced Ebola transmission particularly during the initial phases of the epidemic, resulting in flatter epidemic curves.

CONCLUSIONS

The IBM showed that preventive vaccination may reduce Ebola cases, hospitalizations, and deaths, thus safeguarding the healthcare system and providing more time to implement additional interventions during an outbreak.

Environmental determinants predicting population vulnerability to high yellow fever incidence

Yellow fever (YF) is an endemic mosquito-borne disease in Brazil, though many locations have not observed cases in recent decades. Some locations with low disease burden may resemble locations with higher disease burden through environmental and ecohydrological characteristics, which are known to impact YF burden, motivating increased or continued prevention measures such as vaccination, mosquito control or surveillance. This study aimed to use environmental characteristics to estimate vulnerability to observing high YF burden among all Brazilian municipalities. Vulnerability was defined in three categories based on yearly incidence between 2000 and 2017: minimal, low and high vulnerability. A cumulative logit model was fit to these categories using environmental and ecohydrological predictors, selecting those that provided the most accurate model fit. Per cent of days with precipitation, mean temperature, biome, population density, elevation, vegetation and nearby disease occurrence were included in best-fitting models. Model results were applied to estimate vulnerability nationwide. Municipalities with highest probability of observing high vulnerability was found in the North and Central-West (2000-2016) as well as the Southeast (2017) regions. Results of this study serve to identify specific locations to prioritize new or ongoing surveillance and prevention of YF based on underlying ecohydrological conditions.

Estimating the effect of non-pharmaceutical interventions to mitigate COVID-19 spread in Saudi Arabia

BACKGROUND

The Kingdom of Saudi Arabia (KSA) quickly controlled the spread of SARS-CoV-2 by implementing several non-pharmaceutical interventions (NPIs), including suspension of international and national travel, local curfews, closing public spaces (i.e., schools and universities, malls and shops), and limiting religious gatherings. The KSA also mandated all citizens to respect physical distancing and to wear face masks. However, after relaxing some restrictions during June 2020, the KSA is now planning a strategy that could allow resuming in-person education and international travel. The aim of our study was to evaluate the effect of NPIs on the spread of the COVID-19 and test strategies to open schools and resume international travel.

METHODS

We built a spatial-explicit individual-based model to represent the whole KSA population (IBM-KSA). The IBM-KSA was parameterized using country demographic, remote sensing, and epidemiological data. A social network was created to represent contact heterogeneity and interaction among age groups of the population. The IBM-KSA also simulated the movement of people across the country based on a gravity model. We used the IBM-KSA to evaluate the effect of different NPIs adopted by the KSA (physical distancing, mask-wearing, and contact tracing) and to forecast the impact of strategies to open schools and resume international travels.

RESULTS

The IBM-KSA results scenarios showed the high effectiveness of mask-wearing, physical distancing, and contact tracing in controlling the spread of the disease. Without NPIs, the KSA could have reported 4,824,065 (95% CI: 3,673,775-6,335,423) cases by June 2021. The IBM-KSA showed that mandatory mask-wearing and physical distancing saved 39,452 lives (95% CI: 26,641-44,494). In-person education without personal protection during teaching would have resulted in a high surge of COVID-19 cases. Compared to scenarios with no personal protection, enforcing mask-wearing and physical distancing in schools reduced cases, hospitalizations, and deaths by 25% and 50%, when adherence to these NPIs was set to 50% and 70%, respectively. The IBM-KSA also showed that a quarantine imposed on international travelers reduced the probability of outbreaks in the country.

CONCLUSIONS

This study showed that the interventions adopted by the KSA were able to control the spread of SARS-CoV-2 in the absence of a vaccine. In-person education should be resumed only if NPIs could be applied in schools and universities. International travel can be resumed but with strict quarantine rules. The KSA needs to keep strict NPIs in place until a high fraction of the population is vaccinated in order to reduce hospitalizations and deaths.

Transcriptional Regulation of Carboxylesterase 1 in Human Liver: Role of the Nuclear Receptor Subfamily 1 Group H Member 3 and Its Splice Isoforms

Carboxylesterase 1 (CES1) is the predominant carboxylesterase in the human liver, involved in metabolism of both xenobiotics and endogenous substrates. Genetic or epigenetic factors that alter CES1 activity or expression are associated with changes in drug response, lipid, and glucose homeostasis. However, the transcriptional regulation of CES1 in the human liver remains uncertain. By applying both the random forest and Sobol's Sensitivity Indices (SSI) to analyze existing liver RNA expression microarray data (GSE9588), we identified nuclear receptor subfamily 1 group H member 3 (NR1H3) liver X receptor (LXR)α as a key factor regulating constitutive CES1 expression. This model prediction was validated using small interfering RNA (siRNA) knockdown and CRISPR-mediated transcriptional activation of NR1H3 in Huh7 and HepG2 cells. We found that NR1H3's activation of CES1 is splice isoform-specific, namely that increased expression of the NR1H3-211 isoform increased CES1 expression whereas NR1H3-201 did not. Also, in human liver samples, expression of NR1H3-211 and CES1 are correlated, whereas NR1H3-201 and CES1 are not. This trend also occurs during differentiation of induced pluripotent stem cells (iPSCs) to hepatocytes, where only expression of the NR1H3-211 isoform parallels expression of CES1 Moreover, we found that treatment with the NR1H3 agonist T0901317 in HepG2 cells had no effect on CES1 expression. Overall, our results demonstrate a key role of NR1H3 in maintaining the constitutive expression of CES1 in the human liver. Furthermore, our results support that the effect of NR1H3 is splice isoform-specific and appears to be ligand independent. SIGNIFICANCE STATEMENT: Despite the central role of carboxylesterase 1 (CES1) in metabolism of numerous medications, little is known about its transcriptional regulation. This study identifies nuclear receptor subfamily 1 group H member 3 as a key regulator of constitutive CES1 expression and therefore is a potential target for future studies to understand interperson variabilities in CES1 activity and drug metabolism.

High dimensional model representation of log likelihood ratio: binary classification with SNP data

Background

Developing binary classification rules based on SNP observations has been a major challenge for many modern bioinformatics applications, e.g., predicting risk of future disease events in complex conditions such as cancer. Small-sample, high-dimensional nature of SNP data, weak effect of each SNP on the outcome, and highly non-linear SNP interactions are several key factors complicating the analysis. Additionally, SNPs take a finite number of values which may be best understood as ordinal or categorical variables, but are treated as continuous ones by many algorithms.

Methods

We use the theory of high dimensional model representation (HDMR) to build appropriate low dimensional glass-box models, allowing us to account for the effects of feature interactions. We compute the second order HDMR expansion of the log-likelihood ratio to account for the effects of single SNPs and their pairwise interactions. We propose a regression based approach, called linear approximation for block second order HDMR expansion of categorical observations (LABS-HDMR-CO), to approximate the HDMR coefficients. We show how HDMR can be used to detect pairwise SNP interactions, and propose the fixed pattern test (FPT) to identify statistically significant pairwise interactions.

Results

We apply LABS-HDMR-CO and FPT to synthetically generated HAPGEN2 data as well as to two GWAS cancer datasets. In these examples LABS-HDMR-CO enjoys superior accuracy compared with several algorithms used for SNP classification, while also taking pairwise interactions into account. FPT declares very few significant interactions in the small sample GWAS datasets when bounding false discovery rate (FDR) by 5%, due to the large number of tests performed. On the other hand, LABS-HDMR-CO utilizes a large number of SNP pairs to improve its prediction accuracy. In the larger HAPGEN2 dataset FTP declares a larger portion of SNP pairs used by LABS-HDMR-CO as significant.

Conclusion

LABS-HDMR-CO and FPT are interesting methods to design prediction rules and detect pairwise feature interactions for SNP data. Reliably detecting pairwise SNP interactions and taking advantage of potential interactions to improve prediction accuracy are two different objectives addressed by these methods. While the large number of potential SNP interactions may result in low power of detection, potentially interacting SNP pairs, of which many might be false alarms, can still be used to improve prediction accuracy.

High dimensional model representation of log-likelihood ratio: binary classification with expression data

Background

Binary classification rules based on a small-sample of high-dimensional data (for instance, gene expression data) are ubiquitous in modern bioinformatics. Constructing such classifiers is challenging due to (a) the complex nature of underlying biological traits, such as gene interactions, and (b) the need for highly interpretable glass-box models. We use the theory of high dimensional model representation (HDMR) to build interpretable low dimensional approximations of the log-likelihood ratio accounting for the effects of each individual gene as well as gene-gene interactions. We propose two algorithms approximating the second order HDMR expansion, and a hypothesis test based on the HDMR formulation to identify significantly dysregulated pairwise interactions. The theory is seen as flexible and requiring only a mild set of assumptions.

Results

We apply our approach to gene expression data from both synthetic and real (breast and lung cancer) datasets comparing it also against several popular state-of-the-art methods. The analyses suggest the proposed algorithms can be used to obtain interpretable prediction rules with high prediction accuracies and to successfully extract significantly dysregulated gene-gene interactions from the data. They also compare favorably against their competitors across multiple synthetic data scenarios.

Conclusion

The proposed HDMR-based approach appears to produce a reliable classifier that additionally allows one to describe how individual genes or gene-gene interactions affect classification decisions. Both real and synthetic data analyses suggest that our methods can be used to identify gene networks with dysregulated pairwise interactions, and are therefore appropriate for differential networks analysis.

Explaining income-related inequalities in cardiovascular risk factors in Tunisian adults during the last decade: comparison of sensitivity analysis of logistic regression and Wagstaff decomposition analysis

BACKGROUND

It is important to quantify inequality, explain the contribution of underlying social determinants and to provide evidence to guide health policy. The aim of the study is to explain the income-related inequalities in cardiovascular risk factors in the last decade among Tunisian adults aged between 35 and 70 years old.

METHODS

We performed the analysis by applying two approaches and compared the results provided by the two methods. The methods were global sensitivity analysis (GSA) using logistic regression models and the Wagstaff decomposition analysis.

RESULTS

Results provided by the two methods found a higher risk of cardiovascular diseases and diabetes in those with high socio-economic status in 2005. Similar results were observed in 2016. In 2016, the GSA showed that education level occupied the first place on the explanatory list of factors explaining 36.1% of the adult social inequality in high cardiovascular risk, followed by the area of residence (26.2%) and income (15.1%). Based on the Wagstaff decomposition analysis, the area of residence occupied the first place and explained 40.3% followed by income and education level explaining 19.2 and 14.0% respectively. Thus, both methods found similar factors explaining inequalities (income, educational level and regional conditions) but with different rankings of importance.

CONCLUSIONS

The present study showed substantial income-related inequalities in cardiovascular risk factors and diabetes in Tunisia and provided explanations for this. Results based on two different methods similarly showed that structural disparities on income, educational level and regional conditions should be addressed in order to reduce inequalities.

Ligand-Free Estrogen Receptor α (ESR1) as Master Regulator for the Expression of CYP3A4 and Other Cytochrome P450 Enzymes in the Human Liver

Cytochrome P450 3A4 isoform (CYP3A4) transcription is controlled by hepatic transcription factors (TFs), but how TFs dynamically interact remains uncertain. We hypothesize that several TFs form a regulatory network with nonlinear, dynamic, and hierarchical interactions. To resolve complex interactions, we have applied a computational approach for estimating Sobol's sensitivity indices (SSI) under generalized linear models to existing liver RNA expression microarray data (GSE9588) and RNA-seq data from genotype-tissue expression (GTEx), generating robust importance ranking of TF effects and interactions. The SSI-based analysis identified TFs and interacting TF pairs, triplets, and quadruplets involved in CYP3A4 expression. In addition to known CYP3A4 TFs, estrogen receptor α (ESR1) emerges as key TF with the strongest main effect and as the most frequently included TF interacting partner. Model predictions were validated using small interfering RNA (siRNA)/short hairpin RNA (shRNA) gene knockdown and clustered regularly interspaced short palindromic repeats (CRISPR)-mediated transcriptional activation of ESR1 in biliary epithelial Huh7 cells and human hepatocytes in the absence of estrogen. Moreover, ESR1 and known CYP3A4 TFs mutually regulate each other. Detectable in both male and female hepatocytes without added estrogen, the results demonstrate a role for unliganded ESR1 in CYP3A4 expression consistent with unliganded ESR1 signaling reported in other cell types. Added estrogen further enhances ESR1 effects. We propose a hierarchical regulatory network for CYP3A4 expression directed by ESR1 through self-regulation, cross regulation, and TF-TF interactions. We also demonstrate that ESR1 regulates the expression of other P450 enzymes, suggesting broad influence of ESR1 on xenobiotics metabolism in human liver. Further studies are required to understand the mechanisms underlying role of ESR1 in P450 regulation. SIGNIFICANCE STATEMENT: This study focuses on identifying key transcription factors and regulatory networks for CYP3A4, the main drug metabolizing enzymes in liver. We applied a new computational approach (Sobol's sensitivity analysis) to existing hepatic gene expression data to determine the role of transcription factors in regulating CYP3A4 expression, and used molecular genetics methods (siRNA/shRNA gene knockdown and CRISPR-mediated transcriptional activation) to test these interactions in life cells. This approach reveals a robust network of TFs, including their putative interactions and the relative impact of each interaction. We find that ESR1 serves as a key transcription factor function in regulating CYP3A4, and it appears to be acting at least in part in a ligand-free fashion.