Browsing by Author "Chekouo, Thierry"

Now showing 1 - 4 of 4
  • Thumbnail Image
    ItemOpen Access
    A Bayesian Variable Selection Model for Semi-Continuous Response Using Gaussian Process
    (2023-09-06) Lipman, Danika; Chekouo, Thierry; Deardon, Rob; Wu, Jingjing; Lu, Xuewen; Safo, Sandra; Chekouo, Thierry; Deardon, Rob
    To my knowledge, there is not a statistical method that can perform Bayesian variable selection in a setting where there is a semi-continuous response with a non-linear relationship to predictor variables. I have developed a two-part model to accommodate a semi-continuous response, that uses Gaussian processes to capture the non-linear relationship between input variables and outcomes. Bayesian variable selection is induced in both parts of the model through the construction of the kernel matrices. I have employed the Nystr\"{o}m approximation for kernel matrices to reduce the computational complexity that occurs when working with kernel matrices and large sample sizes. I perform simulation studies and determine my method is competitive in prediction and variable selection with methods such as elastic net, and other methods that capture non-linearity such as random forests, and gradient boosted trees. In addition, I apply my method to a coronary artery disease (CAD) dataset from the Duke Database for Cardiovascular Disease (DDCD) to determine key gene expression features associated with the CAD index, a measure of CAD severity.
  • Thumbnail Image
    ItemOpen Access
    Bayesian Variable Selection Model with Semicontinuous Response
    (2022-01-14) Babatunde, Samuel; Chekouo, Thierry; Sajobi, Tolulope; Zhang, Qingrun; Deardon, Robert; Bezdek, Karoly
    We propose a novel Bayesian variable selection approach that identifies a set of features associated with a semicontinuous response. We used a two-part model where one of the models is a logit model that estimates the probability of zero responses while the other model is a log-normal model that estimates responses greater than zero (positive values). Stochastic Search Variable Selection (SSVS) procedure is used to randomly sample the indicator variables for variable selection which in turn searches the space of feature subsets and identifies the most promising features in the model. For the logistic model, a data augmentation approach is used to sample from the posterior density. We impose a spike-and-slab prior for the regression effects where the unselected covariates take on a prior mass at zero while the selected covariates follow a normal distribution (including the intercept and clinical covariates). Since the joint posterior density had no closed form, we employed the techniques of the Markov Chain Monte Carlo (MCMC) to sample from the posterior distribution. Simulation studies are used to assess the performance of the proposed method. We computed the average area under the receiver operating characteristic curve (AUC) to assess variable selection and compared it with competing methods. We also assessed the convergence diagnosis of our MCMC algorithm by computing the potential scale reduction factor and correlations between the marginal posterior probabilities. We finally apply our method to the coronary artery disease (CAD) data where the aim is to select important genes associated with the CAD index. This data consists of clinical covariates and gene expressions.
  • Thumbnail Image
    ItemOpen Access
    Dementia risk prediction in individuals with mild cognitive impairment: a comparison of Cox regression and machine learning models
    (2022-11-02) Wang, Meng; Greenberg, Matthew; Forkert, Nils D.; Chekouo, Thierry; Afriyie, Gabriel; Ismail, Zahinoor; Smith, Eric E.; Sajobi, Tolulope T.
    Abstract Background Cox proportional hazards regression models and machine learning models are widely used for predicting the risk of dementia. Existing comparisons of these models have mostly been based on empirical datasets and have yielded mixed results. This study examines the accuracy of various machine learning and of the Cox regression models for predicting time-to-event outcomes using Monte Carlo simulation in people with mild cognitive impairment (MCI). Methods The predictive accuracy of nine time-to-event regression and machine learning models were investigated. These models include Cox regression, penalized Cox regression (with Ridge, LASSO, and elastic net penalties), survival trees, random survival forests, survival support vector machines, artificial neural networks, and extreme gradient boosting. Simulation data were generated using study design and data characteristics of a clinical registry and a large community-based registry of patients with MCI. The predictive performance of these models was evaluated based on three-fold cross-validation via Harrell’s concordance index (c-index), integrated calibration index (ICI), and integrated brier score (IBS). Results Cox regression and machine learning model had comparable predictive accuracy across three different performance metrics and data-analytic conditions. The estimated c-index values for Cox regression, random survival forests, and extreme gradient boosting were 0.70, 0.69 and 0.70, respectively, when the data were generated from a Cox regression model in a large sample-size conditions. In contrast, the estimated c-index values for these models were 0.64, 0.64, and 0.65 when the data were generated from a random survival forest in a large sample size conditions. Both Cox regression and random survival forest had the lowest ICI values (0.12 for a large sample size and 0.18 for a small sample size) among all the investigated models regardless of sample size and data generating model. Conclusion Cox regression models have comparable, and sometimes better predictive performance, than more complex machine learning models. We recommend that the choice among these models should be guided by important considerations for research hypotheses, model interpretability, and type of data.
  • Thumbnail Image
    ItemOpen Access
    Integrative multi-omics approach for identifying molecular signatures and pathways and deriving and validating molecular scores for COVID-19 severity and status
    (2023-06-12) Lipman, Danika; Safo, Sandra E.; Chekouo, Thierry
    Abstract Background There is still more to learn about the pathobiology of COVID-19. A multi-omic approach offers a holistic view to better understand the mechanisms of COVID-19. We used state-of-the-art statistical learning methods to integrate genomics, metabolomics, proteomics, and lipidomics data obtained from 123 patients experiencing COVID-19 or COVID-19-like symptoms for the purpose of identifying molecular signatures and corresponding pathways associated with the disease. Results We constructed and validated molecular scores and evaluated their utility beyond clinical factors known to impact disease status and severity. We identified inflammation- and immune response-related pathways, and other pathways, providing insights into possible consequences of the disease. Conclusions The molecular scores we derived were strongly associated with disease status and severity and can be used to identify individuals at a higher risk for developing severe disease. These findings have the potential to provide further, and needed, insights into why certain individuals develop worse outcomes.