Zero-inflated Negative Binomial Research Articles

MODELING THE INCIDENCE OF MALNUTRITION IN BOGOR REGENCY USING ZERO-INFLATED NEGATIVE BINOMIAL MIXED EFFECT MODEL

Modeling response variables in the form of count data generally uses a model based on the Poisson distribution. However, some conditions, such as the presence of excess zero, can be found in the data that result in overdispersion, which will have an impact on the resulting variance in the model. In this paper, three approaches, namely the Poisson Mixed Model, the Negative Binomial (NB) Mixed Model, and the Zero-Inflated Negative Binomial (ZINB) Mixed Model, are used to model the incidence of malnutrition in Bogor Regency. The data used in this study are secondary data sourced from the West Java open data website. Based on the results of data analysis, it appears that the ZINB Mixed Model method is a method capable of accommodating random effects, overdispersion, and excess zero in modeling malnutrition in Bogor Regency. Variables that significantly affect the occurrence of malnutrition cases in villages in Bogor Regency include the Number of Children Weighed Routinely Every Month, Number of Children Measured for Length and Height Twice a Year, Number of Children under 12 Months Old Who Received Complete Basic Immunization, Number of Posyandu (Integrated Health Post), and Number of Parents/Caregivers Participating in Monthly Parenting (PAUD).

Factors affecting firm‐level job cuts during the COVID‐19 pandemic: A cross‐country evidence

AbstractThis study investigates the determinants of firms' job‐cut decisions during the COVID‐19 pandemic, considering both firm‐level and country‐level factors. Data from 31 countries (a mix of developed and emerging) collected between May 2020 and May 2021 are analyzed using a multilevel Zero‐Inflated Negative Binomial (ZINB) model. The results reveal that firms that were operational, larger in size, received financial incentives, and arranged remote work for their workforce laid off a smaller proportion of workers. Conversely, firms that experienced significant sales reductions, input supply disruptions, and introduced delivery or carry‐out services laid off a larger proportion of workers. Moreover, among financial incentive‐recipient firms, smaller ones and those that introduced remote work and delivery or carry‐out services had smaller layoffs. At the country level, the human capital index (HCI) significantly influenced job‐cut decisions, with higher HCI scores associated with smaller layoffs. Classifying countries into “developed” and “emerging” yielded similar results, except for temporary closure having no significant impact on job cuts in developed countries and remote work showing no impact on job cuts in emerging countries. The robustness of the results was confirmed by a multilevel zero‐inflated Tobit model, which consistently reproduced the outcomes.

Immigrant background, incarceration history and recidivism among adults assessed for illicit substance use severity: Findings from a Swedish registry study

BackgroundIn Sweden, as in other countries, individuals with immigrant backgrounds are disproportionately represented within the incarcerated population. This study examined the association between immigrant background and future incarceration for individuals assessed for illicit substance use severity, while considering their prior incarceration history. MethodsUsing data from Swedish Addiction Severity Index (ASI) assessments linked to register data from Statistics Sweden, we employed Zero-Inflated Negative Binomial (ZINB) regression models to analyse differences in incarceration histories within five years before and after ASI assessments based on immigrant backgrounds. Additionally, Cox proportional-hazard models were used to assess the likelihood of post-assessment incarceration among these groups. ResultsImmigrant background was positively associated with pre- and post-assessment incarceration. First- and second-generation immigrants from the Global South had longer periods of incarceration in the five years before assessments compared to native Swedes. Post-assessment, first-generation immigrants showed longer periods of incarceration. Survival analyses supported these findings, indicating a higher risk of prolonged post-assessment incarceration among all immigrant groups, particularly first-generation immigrants from the Global South. ConclusionAmong individuals assessed for illicit drug use within Swedish municipalities, those with immigrant backgrounds faced higher incarceration risks, even after controlling for substance use severity and prior incarceration. Tailored interventions and support systems are vital to prevent re-entry into the criminal justice system. Timely actions can break re-offending cycles, redirecting paths away from reoffending and towards legal reintegration, thereby reducing incarceration and recidivism rates.

Unsupervised Gene-Cell Collective Representation Learning with Optimal Transport

Cell type identification plays a vital role in single-cell RNA sequencing (scRNA-seq) data analysis. Although many deep embedded methods to cluster scRNA-seq data have been proposed, they still fail in elucidating the intrinsic properties of cells and genes. Here, we present a novel end-to-end deep graph clustering model for single-cell transcriptomics data based on unsupervised Gene-Cell Collective representation learning and Optimal Transport (scGCOT) which integrates both cell and gene correlations. Specifically, scGCOT learns the latent embedding of cells and genes simultaneously and reconstructs the cell graph, the gene graph, and the gene expression count matrix. A zero-inflated negative binomial (ZINB) model is estimated via the reconstructed count matrix to capture the essential properties of scRNA-seq data. By leveraging the optimal transport-based joint representation alignment, scGCOT learns the clustering process and the latent representations through a mutually supervised self optimization strategy. Extensive experiments with 14 competing methods on 15 real scRNA-seq datasets demonstrate the competitive edges of scGCOT.

Environmental dynamics and seasonal wing physiological changes of coastal mosquito vectors (Diptera: Culicidae) in Central Thailand

Mosquitoes (Diptera: Culicidae) are small hematophagous insects that serve as disease vectors by transmitting pathogens to humans and animals. This study was conducted on the coastline of Samut Songkhram Province in Thailand's central region to evaluate the impact of environmental factors on coastal mosquitoes, namely Anopheles epiroticus and Culex sitiens. A zero-inflated negative binomial (ZINB) regression was employed to assess the relationship between the density of larval mosquitoes and water quality factors. Spearman's rank correlation was used to examine the relationship between the number of adult mosquitoes and meteorological factors, and the geometric morphometric (GM) technique was used to study the wing size and shape of adult mosquitoes throughout different seasons. The ZINB analysis revealed a significant negative relationship (p < 0.05) between dissolved oxygen, high-range phosphate, and the number of An. epiroticus larvae. Conversely, in the case of Cx. sitiens, there was a significant positive relationship between nitrate-nitrogen and larval density (p < 0.01), while dissolved oxygen exhibited a significant negative relationship (p < 0.05). Regarding adult mosquitoes, relative humidity was positively correlated with the abundance of adult Cx. sitiens (r = 0.85, p < 0.01), whereas each meteorological parameter—temperature, relative humidity, rainfall, wind speed, and atmospheric pressure—was negatively correlated with adult An. epiroticus abundance (p < 0.05 for each). The GM analysis revealed statistically significant seasonal variations in wing shapes for both species (p < 0.05 for all seasonal comparisons), underscoring their adaptation to seasonal environmental conditions in coastal habitats.

Examining Crime-Specific and Crime-General Theories of Crime Causation at Place: The Case of Property and Violent Crime on Street Segments in Tel Aviv-Yafo

AbstractWe use data on street segments in Tel Aviv-Yafo to examine whether general or specific crime factors are key to understanding crime rates on street segments. We pose two questions. (1) What causal factors explain the property and violent crimes at the same locations? (2) What are the differences and similarities between the risk factors of the two crime types? Our study capitalizes on data drawn from the Israeli Central Bureau of Statistics (CBS) to identify social and opportunity data at the street segment level. The dependent variable is the average counts of violent and property crimes in Tel Aviv-Yafo between 2010 and 2014. Zero-inflated negative binomial (ZINB) regression explains violent and property crime variations. While many of the significant factors that explain crime are similar between the two crime types, there are also distinct criminogenic factors predicting violent and property crime. Overall, our results support the position of common or general crime causes at places, but at the same time, they suggest the importance of understanding specific causes for specific crime types.

Factors Influencing Time to Return to Learn Among NCAA Student-Athletes Enrolled in the Concussion Assessment, Research, and Education (CARE) Study.

The aim was to describe the demographic and post-injury factors that influence time to return to learn (RTL) among student-athletes enrolled in the Concussion Assessment, Research and Education (CARE) Consortium. A total of 47,860 student-athletes enrolled in the National Collegiate Athletic Association-Department of Defense (NCAA-DoD) CARE Consortium study from 2014 to 2020, with 1485 sport-related concussions (SRCs) analyzed in the present dataset. Demographic and post-injury characteristics were calculated using descriptive statistics, followed by Kaplan-Meier estimates to examine median time to return to normal academic performance (i.e., RTL) by sex (male, female), baseline psychiatric conditions (depression, anxiety) and/or learning disorder, NCAA division (I, II, III), SRC history (0, 1, 2, 3+), NCAA sport category (contact, limited contact, non-contact sport), and median difference in baseline/post-injury symptom severity scores (< 21, ≥ 21). Further, a multivariable zero-inflated negative binomial (ZINB) regression model was used to examine their association with RTL. Overall, time to RTL (mean = 4.96 ± 8.24days, median = 3.0 [interquartile range = 0.0, 6.0] days) was found to be influenced by several factors resulting in earlier trajectories. Notably, nearly 25% of the sample demonstrated immediate RTL (i.e., 0days). Among student-athletes who did not immediately RTL, males demonstrated a decreased rate in RTL (rate = 0.79; 95% CI 0.66-0.96) compared to females. Further, student-athletes with a ≥ 21 change in symptom severity score (post-injury baseline) demonstrated a higher rate of RTL (rate = 1.47; 95% CI 1.21-1.79) compared to student-athletes with a symptom severity change score < 21. Lastly, male student-athletes demonstrated two times higher odds (odds ratio = 1.95; 95% CI 1.02-3.73) of immediate RTL compared to female student-athletes. No other covariates were associated with time to RTL. Collectively, the present findings suggest a rapid return to the classroom following concussion. Specifically, males demonstrated higher odds of time to RTL, whereas those with greater differences in symptom severity resulted in a higher rate of time to RTL among those who did not immediately RTL. Ultimately, these findings support prior work emphasizing an individualized approach to SRC management.

Analyzing the Factors Influencing Time Delays in Korean Railroad Accidents

Railroads play a pivotal role in the Korean national economy, necessitating a thorough understanding of factors influencing accidents for effective mitigation strategies. Unlike prior research focused on accident frequency and severity, this study delves into the often-overlooked aspect of time delays resulting from railroad accidents. Analyzing 15 years of nationwide data (2008–2022), encompassing 3244 human-related and 3350 technical events, this research identifies key factors influencing delay likelihood and duration. Factors considered include event type, season, train type, location, operator size, person type involved, facility type, and causes. Despite an overall decrease in events, variable delay times highlight the need to comprehend specific contributing factors. To address excess zeros, the study employs a two-stage model and a zero-inflated negative binomial (ZINB) model, alongside artificial neural networks (ANNs) for non-linear pattern recognition. Human-related delays are influenced by event types, seasons, and passenger categories, exhibit nuanced impacts. Technical-related delays are influenced by incident types and facility involvement. Regarding model performance, the ANN models outperform regression-based models consistently in all cases. This study emphasizes the importance of considering both human and technical factors in predicting and understanding railroad accident delays, offering valuable insights for formulating strategies to mitigate service disruptions associated with these incidents.

How observation of other shoppers increases the in-store use of mobile technology

Digital technologies enable brick-and-mortar retailers to improve customer engagement and retention. In this study, the authors investigate how in-store use of mobile technology can leverage the social nature of offline shopping to increase customer usage of a retailer's promotional mobile app. Analyzing a large dataset from a nationwide retailer by estimating a Zero-Inflated Negative Binomial (ZINB) model, they find a significant effect of in-store observation of other shoppers' use of the retailer's mobile promotion app on both app usage and the number of m-coupon redemptions. However, in-store advertising reduces the effect of observing other shoppers on the number of m-coupons redeemed per app usage.

ZERO-INFLATED NEGATIVE BINOMIAL MODELING IN INFANT DEATH CASE DUE TO PNEUMONIA IN EAST JAVA PROVINCE

Pneumonia is an acute infectious disease of the respiratory tract and an infection caused by a virus, bacteria or fungus that attacks the lung tissue. Several cases of pneumonia have resulted in deaths that occurred in toddlers aged 12-59 months. Based on official health in profile data, East Java's health in 2021 has a zero number of deaths under five aged 12-59 months due to pneumonia. Modeling data with many response variables is zero and there is overdispersion can be done using Zero Inflated Negative Binomial (ZINB) regression. This study aims to model the number of infant deaths aged 12-59 months due to pneumonia in East Java Province based on seven factors that are considered to influence the number of deaths in infants due to pneumonia. From this model, it can be seen that the factors that significantly influence the death of infants aged 12-59 months due to pneumonia in East Java Province using Zero Inflated Negative Binomial (ZINB) regression. The results of testing the parameters of the ZINB regression model show that the predictor variables that have a partial significant effect on the negative binomial model in East Java are the percentage of infants who received complete basic immunization, the percentage of coverage of under-five health services, the percentage of under-five children with malnutrition, the percentage of LBW (low birth weight babies). Selection of the best model is obtained by using the Bayesian Information Criterion (BIC) of 101,587.

Is review visibility fostering helpful votes? The role of review rank and review characteristics in the adoption of information

In online environments, where consumers usually face information overload, information regarding the number of helpful votes received by online reviews serves as a trust sign to aid consumers in their purchasing journeys. As consumers can only vote for a review as helpful if they have viewed it, the position of the review in the sequence of reviews is likely to influence the number of helpful votes that the review receives. We propose a model in which review helpfulness depends not only on the characteristics of the review and reviewer, but also on its visibility. Review visibility is defined in this study as the probability of a review being viewed by a consumer, and is measured by the inverse rank order of the review in the sequence of reviews at the online retailer when consumers sort reviews according to different criteria (most helpful and most recent). Using a database of 59,526 online reviews from a popular cosmetics online store in the US, we estimate a zero-inflated negative binomial (ZINB) regression and find evidence that review visibility has a strong impact in explaining the likelihood of a review being read by consumers and subsequently voted as helpful by consumers. This effect is even stronger when sorting is most helpful.

Predictors of frequency of 1-year readmission in adult patients with diabetes

Diabetes mellitus (DM) is the third most common chronic condition associated with frequent hospital readmissions. Predictors of the number of readmissions within 1 year among patients with DM are less often studied compared with those of 30-day readmission. This study aims to identify predictors of number of readmissions within 1 year amongst adult patients with DM and compare different count regression models with respect to model fit. Data from 2008 to 2015 were extracted from the electronic medical records of the National University Hospital, Singapore. Inpatients aged ≥ 18 years at the time of index admission with a hospital stay > 24 h and survived until discharge were included. The zero-inflated negative binomial (ZINB) model was fitted and compared with three other count models (Poisson, zero-inflated Poisson and negative binomial) in terms of predicted probabilities, misclassification proportions and model fit. Adjusted for other variables in the model, the expected number of readmissions was 1.42 (95% confidence interval [CI] 1.07 to 1.90) for peripheral vascular disease, 1.60 (95% CI 1.34 to 1.92) for renal disease and 2.37 (95% CI 1.67 to 3.35) for Singapore residency. Number of emergency visits, number of drugs and age were other significant predictors, with length of stay fitted as a zero-inflated component. Model comparisons suggested that ZINB provides better prediction than the other three count models. The ZINB model identified five patient characteristics and two comorbidities associated with number of readmissions. It outperformed other count regression models but should be validated before clinical adoption.

Effect of the measurement period and spatial dependence on the accuracy of urban freight trip generation models

Despite recent advancements in the Freight Trip Generation (FTG) modelling literature, there is a lack of understanding on the effect of the choice of a regression model, measurement period (daily/weekly FTG), and spatial dependence on model fit and freight-related policies. This study addresses these research gaps by developing non-spatial and spatial autoregressive multiple linear regression and count models for daily and weekly Freight Trip Production (FTP) and Freight Trip Attraction (FTA). We model Freight Shipments (FS) as FTP and Freight Deliveries (FD) as FTA. The results show that the best model for daily and weekly FTP is the spatial Zero-Inflated Negative Binomial (ZINB) model. The best daily and weekly FTA model is the non-spatial Negative Binomial (NB) model. The findings indicate the presence of spatial dependence in the best FTP model, while it is absent in the best FTA model. The elasticity analysis shows that daily models may lead to bias and inaccurate prediction of policy impacts. The study recommends using count models that capture more FTG characteristics with a week as the measurement period and consider spatial dependence, if present.

Attention-based deep clustering method for scRNA-seq cell type identification.

Single-cell sequencing (scRNA-seq) technology provides higher resolution of cellular differences than bulk RNA sequencing and reveals the heterogeneity in biological research. The analysis of scRNA-seq datasets is premised on the subpopulation assignment. When an appropriate reference is not available, such as specific marker genes and single-cell reference atlas, unsupervised clustering approaches become the predominant option. However, the inherent sparsity and high-dimensionality of scRNA-seq datasets pose specific analytical challenges to traditional clustering methods. Therefore, a various deep learning-based methods have been proposed to address these challenges. As each method improves partially, a comprehensive method needs to be proposed. In this article, we propose a novel scRNA-seq data clustering method named AttentionAE-sc (Attention fusion AutoEncoder for single-cell). Two different scRNA-seq clustering strategies are combined through an attention mechanism, that include zero-inflated negative binomial (ZINB)-based methods dealing with the impact of dropout events and graph autoencoder (GAE)-based methods relying on information from neighbors to guide the dimension reduction. Based on an iterative fusion between denoising and topological embeddings, AttentionAE-sc can easily acquire clustering-friendly cell representations that similar cells are closer in the hidden embedding. Compared with several state-of-art baseline methods, AttentionAE-sc demonstrated excellent clustering performance on 16 real scRNA-seq datasets without the need to specify the number of groups. Additionally, AttentionAE-sc learned improved cell representations and exhibited enhanced stability and robustness. Furthermore, AttentionAE-sc achieved remarkable identification in a breast cancer single-cell atlas dataset and provided valuable insights into the heterogeneity among different cell subtypes.

PSXIV-18 Fit of the Zero-Inflated Negative Binomial Model to Analyze Fecal Egg Counts

Abstract Fecal egg count (FEC) is used as an indicator of parasite infection level in sheep. The distribution of FEC is non-Gaussian and typically overdispersed, often with an excess in zero counts. Quantifying the extent of inflation of zero counts can be difficult. Our objective was to assess the potential zero-inflation problem in FEC resulting from variation in infection with gastrointestinal nematodes by using a generalized linear model approach. The zero-inflated Poisson (ZIP) and zero-inflated negative binomial (ZINB) models are useful techniques to analyze count data with excess zeros; ZINB also handles overdispersion. The ZINB model has the potential to delineate ‘true’ zeros, in this case, animals resistant to parasitism and thereby with zero FEC, from ‘false’ zeros, animals never or minimally exposed to a parasite challenge. By distinguishing false zeros, those animals expressing parasite resistance may be more clearly identified. Two datasets on Katahdin sheep, a hair breed known to express resistance to gastrointestinal nematodes, were investigated; a smaller set (n = 3,048) with FEC and FAMACHA (FAM) scores, a subjective measure of anemia indicative of parasitism by Haemonchus contortus, a blood sucking helminth; and a larger set (n = 14,405) with FEC and a contemporary group (CG) designation, assigned based on the flock, birth year, management group, and FEC recording date of the animal. Among animals with FAM recorded, 14% had scores indicative of at least border line anemia. Amongst the 410 CG, 22% had mean FEC more than 500 egg/g, a threshold value routinely used to indicate a substantial infection level with H. contortus. For each dataset, the Poisson, Negative Binomial, ZIP, and ZINB models were fit and compared using R (pscl package) and SAS software, with the ZINB providing the best fit. In the models considered, FEC was the response variable and either FAM or CG was the explanatory variable, depending on the dataset. Despite evidence of parasite challenge, the true and false zeros could not be delineated in both data sets using these models. The estimated proportion of false zeros due to inflation did not differ from the proportion of zeros observed in the data set. Either all zeros coincided with no infection, which seems unlikely in Katahdins, or neither FAM nor CG provided sufficient information to distinguish resistant from uninfected individuals. Alternative or additional explanatory variables, such as packed cell volume or immunoglobulin concentrations indicative of parasitic infection, may be necessary to separate true from false zero FEC in sheep challenged with gastrointestinal nematodes using the ZINB model.

Geographically Weighted Zero-Inflated Negative Binomial Regression: A general case for count data

Poisson and Negative Binomial Regression Models are often used to describe the relationship between a count dependent variable and a set of independent variables. However, these models fail to analyze data with an excess of zeros, being Zero-Inflated Poisson (ZIP) and Zero-Inflated Negative Binomial (ZINB) models the most appropriate to fit this kind of data. To Incorporate the spatial dimension into the count data models, Geographically Weighted Poisson Regression (GWPR), Geographically Weighted Negative Binomial Regression (GWNBR) and Geographically Weighted Zero-Inflated Poisson Regression (GWZIPR) have been developed, but the zero-inflation part of the negative binomial distribution is undeveloped in order to incorporate the overdispersion and the excess of zeros, as was at the beginning of the COVID-19 pandemic, whereas some places were having an outbreak of cases and in others places, there were no cases yet. Therefore, we propose a Geographically Weighted Zero-Inflated Negative Binomial Regression (GWZINBR) model which can be considered a general case for count data, since locally it can become a GWZIPR, GWNBR or a GWPR model. We applied this model to simulated data and to the cases of COVID-19 in South Korea at the beginning of the pandemic in 2020 and the results showed a better understanding of the phenomenon compared to the GWNBR model.

Investigating Highway–Rail Grade Crossing Inventory Data Quality’s Role in Crash Model Estimation and Crash Prediction

The highway–rail grade crossings (HRGCs) crash frequency models used in the US are based on the Federal Railroad Administration’s (FRA) database for highway–rail crossing inventory. Inaccuracies or missing values within this database directly impact the estimated parameters of the crash models and subsequent crash predictions. Utilizing a set of 560 HRGCs in Nebraska, this research demonstrates variations in crash predictions estimated by the FRA’s 2020 Accident Prediction (AP) model under two scenarios: firstly, employing the unchanged, original FRA HRGCs inventory dataset as the input, and secondly, utilizing a field-validated inventory dataset for the same 560 HRGCs as input to the FRA’s 2020 Accident Prediction (AP) model. The findings indicated a significant statistical disparity in the predictions made with the two input datasets. Furthermore, two new Zero-inflated Negative Binomial (ZINB) models were estimated by employing 5-year reported HRGCs crashes and the two inventory datasets for the 560 HRGCs. These models facilitated the comparison of model parameter estimates and estimated marginal values. The results indicated that errors and missing values in the original FRA HRGCs inventory dataset resulted in crash predictions that statistically differed from those made using the more accurate and complete (validated in the field) HRGCs inventory dataset. Furthermore, the crash prediction model estimated upon the corrected inventory data demonstrated enhanced prediction performance, as measured by the statistical fitness criteria. The findings emphasize the importance of collecting complete and accurate inventory data when developing HRGCs crash frequency models. This will enhance models’ precision, improve their predictive capabilities to aid in better resource allocation, and ultimately reduce HRGCs crashes.

Photosynthetic animals and where to find them: abundance and size of a solar-powered sea slug in different light conditions

Three populations of the solar-powered sea slug Elysia crispata on reefs from the southern Gulf of Mexico and the Caribbean were analyzed. The aim was to describe and compare the changes in abundance and size of this species in different localities, as a function of depth and time-of-day. We hypothesized that differences in abundance would be related to locality, time of the day and depth, and differences in size would be related to locality and time of the day. Using snorkeling and SCUBA diving, all individuals within quadrats were counted and measured. A total of 680 organisms were recorded at Verde, Arcas and Puerto Morelos (PM) reefs at five times of the day (sunrise, morning, zenith, evening, and night) and depths of 0–13 m. Zero inflated negative binomial (ZINB) regressions adjusted to abundance data showed that E. crispata in Arcas and Verde reefs is expected to be more abundant (> 50) in shallow depths (< 2 m) at any time of the day except sunrise, whereas a low abundance (≤ 1 organism) is predicted in PM regardless of depth and time-of-day. According to linear models, size was not related to depth, but was related to locality and time-of-day, with sea slugs from Arcas and Verde having similar size, and both larger than those in PM. This information suggests that this sea slug is capable of moving within the reefs and helps to understand the unique biological phenomena of kleptoplasty.

Count regression models analysis of factors affecting antenatal care utilization in Ethiopia: original article.

Antenatal care (ANC) reduces maternal and newborn mortalities and makes it easier to find infections early and prevent them from progressing. This study aimed to identify factors affecting ANC utilization in Ethiopia via the count regression model. The data for the study was drawn from the 2019 Ethiopian Mini Demographic and Health Survey dataset. Count regression models, such as Poisson, Negative Binomial (NB), Zero-Inflated Poisson (ZIP), and Zero-Inflated Negative Binomial (ZINB) models, were used to determine the factors influencing ANC utilization. A total of 3962 women were included in the study. Only about 42% of women used the WHO-recommended number of ANC of a minimum of four visits. The ZIP model was outperforming to fit the data as compared to other count regression models. Rural residence (IRR=0.8832, 95% CI: 0.8264-0.9440), being resident of the Somalia region (IRR=0.4762, 95% CI: 0.3919-0.5785), SNNPR (IRR=0.8716, 95% CI: 0.7913-0.9600), and Gambela Region (IRR=0.7830, 95% CI: 0.7063-0.8680), being Muslim (IRR=0.9384, 95% CI: 0.8876-0.9921) decrease the ANC utilization. Contrarily, Addis Ababa residence (IRR=1.1171, 95% CI: 1.0181-1.2259), primary education (IRR=1.1278, 95% CI: 1.0728-1.1855), secondary and higher education (IRR=1.2357, 95% CI: 1.1550-1.3220), middle wealth index (IRR=1.0855, 95% CI: 1.0095-1.1671) and rich wealth index (IRR=1.0941, 95% CI: 1.0152-1.1790) increase the ANC utilization. The ZIP model best fitted the data compared to others. The study revealed that being poor, rural resident, uneducated, Somalia region resident, SNNPR resident, Gambela region resident, and Muslim were factors associated with lower ANC utilization. Thus, health education is needed to be given for mothers with no education. In addition, building a strong awareness-creation program regarding ANC is required for rural residents so as to improve the level of ANC utilization in Ethiopia.

STGNNks: Identifying cell types in spatial transcriptomics data based on graph neural network, denoising auto-encoder, and [formula omitted]-sums clustering

Background:Spatial transcriptomics technologies fully utilize spatial location information, tissue morphological features, and transcriptional profiles. Integrating these data can greatly advance our understanding about cell biology in the morphological background. Methods:We developed an innovative spatial clustering method called STGNNks by combining graph neural network, denoising auto-encoder, and k-sums clustering. First, spatial resolved transcriptomics data are preprocessed and a hybrid adjacency matrix is constructed. Next, gene expressions and spatial context are integrated to learn spots’ embedding features by a deep graph infomax-based graph convolutional network. Third, the learned features are mapped to a low-dimensional space through a zero-inflated negative binomial (ZINB)-based denoising auto-encoder. Fourth, a k-sums clustering algorithm is developed to identify spatial domains by combining k-means clustering and the ratio-cut clustering algorithms. Finally, it implements spatial trajectory inference, spatially variable gene identification, and differentially expressed gene detection based on the pseudo-space-time method on six 10x Genomics Visium datasets. Results:We compared our proposed STGNNks method with five other spatial clustering methods, CCST, Seurat, stLearn, Scanpy and SEDR. For the first time, four internal indicators in the area of machine learning, that is, silhouette coefficient, the Davies-Bouldin index, the Caliniski-Harabasz index, and the S_Dbw index, were used to measure the clustering performance of STGNNks with CCST, Seurat, stLearn, Scanpy and SEDR on five spatial transcriptomics datasets without labels (i.e., Adult Mouse Brain (FFPE), Adult Mouse Kidney (FFPE), Human Breast Cancer (Block A Section 2), Human Breast Cancer (FFPE), and Human Lymph Node). And two external indicators including adjusted Rand index (ARI) and normalized mutual information (NMI) were applied to evaluate the performance of the above six methods on Human Breast Cancer (Block A Section 1) with real labels. The comparison experiments elucidated that STGNNks obtained the smallest Davies-Bouldin and S_Dbw values and the largest Silhouette Coefficient, Caliniski-Harabasz, ARI and NMI, significantly outperforming the above five spatial transcriptomics analysis algorithms. Furthermore, we detected the top six spatially variable genes and the top five differentially expressed genes in each cluster on the above five unlabeled datasets. And the pseudo-space-time tree plot with hierarchical layout demonstrated a flow of Human Breast Cancer (Block A Section 1) progress in three clades branching from three invasive ductal carcinoma regions to multiple ductal carcinoma in situ sub-clusters. Conclusion:We anticipate that STGNNks can efficiently improve spatial transcriptomics data analysis and further boost the diagnosis and therapy of related diseases. The codes are publicly available at https://github.com/plhhnu/STGNNks.