arXiv subjects

Giovanni Parmigiani

Publications and source records attributed to Giovanni Parmigiani.

50 records · Page 3Linked to original sources

arXiv2019

Using Machine Learning and Natural Language Processing to Review and Classify the Medical Literature on Cancer Susceptibility Genes

PURPOSE: The medical literature relevant to germline genetics is growing exponentially. Clinicians need tools monitoring and prioritizing the literature to understand the clinical implications of the pathogenic genetic variants. We developed and evaluated two machine learning models to classify abstracts as relevant to the penetrance (risk of cancer for germline mutation carriers) or prevalence of germline genetic mutations. METHODS: We conducted literature searches in PubMed and retrieved paper titles and abstracts to create an annotated dataset for training and evaluating the two machine learning classification models. Our first model is a support vector machine (SVM) which learns a linear decision rule based on the bag-of-ngrams representation of each title and abstract. Our second model is a convolutional neural network (CNN) which learns a complex nonlinear decision rule based on the raw title and abstract. We evaluated the performance of the two models on the classification of papers as relevant to penetrance or prevalence. RESULTS: For penetrance classification, we annotated 3740 paper titles and abstracts and used 60% for training the model, 20% for tuning the model, and 20% for evaluating the model. The SVM model achieves 89.53% accuracy (percentage of papers that were correctly classified) while the CNN model achieves 88.95 % accuracy. For prevalence classification, we annotated 3753 paper titles and abstracts. The SVM model achieves 89.14% accuracy while the CNN model achieves 89.13 % accuracy. CONCLUSION: Our models achieve high accuracy in classifying abstracts as relevant to penetrance or prevalence. By facilitating literature review, this tool could help clinicians and researchers keep abreast of the burgeoning knowledge of gene-cancer associations and keep the knowledge bases for clinical decision support tools up to date.

cs.IR↗

arXiv2019

The Fuzzy ROC

The fuzzy ROC extends Receiver Operating Curve (ROC) visualization to the situation where some data points, falling in an indeterminacy region, are not classified. It addresses two challenges: definition of sensitivity and specificity bounds under indeterminacy; and visual summarization of the large number of possibilities arising from different choices of indeterminacy zones.

stat.OT↗

arXiv2018

High-dimensional confounding adjustment using continuous spike and slab priors

In observational studies, estimation of a causal effect of a treatment on an outcome relies on proper adjustment for confounding. If the number of the potential confounders ($p$) is larger than the number of observations ($n$), then direct control for all potential confounders is infeasible. Existing approaches for dimension reduction and penalization are generally aimed at predicting the outcome, and are less suited for estimation of causal effects. Under standard penalization approaches (e.g. Lasso), if a variable $X_j$ is strongly associated with the treatment $T$ but weakly with the outcome $Y$, the coefficient $β_j$ will be shrunk towards zero thus leading to confounding bias. Under the assumption of a linear model for the outcome and sparsity, we propose continuous spike and slab priors on the regression coefficients $β_j$ corresponding to the potential confounders $X_j$. Specifically, we introduce a prior distribution that does not heavily shrink to zero the coefficients ($β_j$s) of the $X_j$s that are strongly associated with $T$ but weakly associated with $Y$. We compare our proposed approach to several state of the art methods proposed in the literature. Our proposed approach has the following features: 1) it reduces confounding bias in high dimensional settings; 2) it shrinks towards zero coefficients of instrumental variables; and 3) it achieves good coverages even in small sample sizes. We apply our approach to the National Health and Nutrition Examination Survey (NHANES) data to estimate the causal effects of persistent pesticide exposure on triglyceride levels.

stat.ME↗

arXiv2018

Multi-study Factor Analysis

We introduce a novel class of factor analysis methodologies for the joint analysis of multiple studies. The goal is to separately identify and estimate 1) common factors shared across multiple studies, and 2) study-specific factors. We develop a fast Expectation Conditional-Maximization algorithm for parameter estimates and we provide a procedure for choosing the common and specific factor. We present simulations evaluating the performance of the method and we illustrate it by applying it to gene expression data in ovarian cancer. In both cases, we clarify the benefits of a joint analysis compared to the standard factor analysis. We hope to have provided a valuable tool to accelerate the pace at which we can combine unsupervised analysis across multiple studies, and understand the cross-study reproducibility of signal in multivariate data.

stat.AP↗

arXiv2018

Bayesian Multi-study Factor Analysis for High-throughput Biological Data

This paper presents a new modeling strategy for joint unsupervised analysis of multiple high-throughput biological studies. As in Multi-study Factor Analysis, our goals are to identify both common factors shared across studies and study-specific factors. Our approach is motivated by the growing body of high-throughput studies in biomedical research, as exemplified by the comprehensive set of expression data on breast tumors considered in our case study. To handle high-dimensional studies, we extend Multi-study Factor Analysis using a Bayesian approach that imposes sparsity. Specifically, we generalize the sparse Bayesian infinite factor model to multiple studies. We also devise novel solutions for the identification of the loading matrices: we recover the loading matrices of interest ex-post, by adapting the orthogonal Procrustes approach. Computationally, we propose an efficient and fast Gibbs sampling approach. Through an extensive simulation analysis, we show that the proposed approach performs very well in a range of different scenarios, and outperforms standard Factor analysis in all the scenarios identifying replicable signal in unsupervised genomic applications. The results of our analysis of breast cancer gene expression across seven studies identified replicable gene patterns, clearly related to well-known breast cancer pathways. An R package is implemented and available on GitHub.

stat.AP↗

arXiv2016

Logistic push: a regression framework for partial AUC optimization

The area under the receiver operating characteristic curve (AUC) is often used to evaluate the performance of clinical prediction models. Recently, a more refined strategy has been proposed to examine a partial area under the curve (pAUC), which can account for differing costs associated with false negative versus false positive results. Such consideration can substantially increase the clinical utility of prediction models depending on the clinical question. Properties of the pAUC estimator create significant challenges for pAUC-optimal marker selection and model building. As such, current approaches towards these aims can be complex and computationally intensive. We present a simpler method based on weighted logistic regressions. We refer to our strategy as logistic push, due to shared heuristics with the ranking algorithm P-norm push. Logistic push is particularly useful in the high-dimensional setting, where fast and broadly available algorithms for fitting penalized regressions can be used for both marker selection and model fitting.

stat.AP↗

arXiv2015

Generalized Quantile Treatment Effect: A Flexible Bayesian Approach Using Quantile Ratio Smoothing

We propose a new general approach for estimating the effect of a binary treatment on a continuous and potentially highly skewed response variable, the generalized quantile treatment effect (GQTE). The GQTE is defined as the difference between a function of the quantiles under the two treatment conditions. As such, it represents a generalization over the standard approaches typically used for estimating a treatment effect (i.e., the average treatment effect and the quantile treatment effect) because it allows the comparison of any arbitrary characteristic of the outcome's distribution under the two treatments. Following Dominici et al. (2005), we assume that a pre-specified transformation of the two quantiles is modeled as a smooth function of the percentiles. This assumption allows us to link the two quantile functions and thus to borrow information from one distribution to the other. The main theoretical contribution we provide is the analytical derivation of a closed form expression for the likelihood of the model. Exploiting this result we propose a novel Bayesian inferential methodology for the GQTE. We show some finite sample properties of our approach through a simulation study which confirms that in some cases it performs better than other nonparametric methods. As an illustration we finally apply our methodology to the 1987 National Medicare Expenditure Survey data to estimate the difference in the single hospitalization medical cost distributions between cases (i.e., subjects affected by smoking attributable diseases) and controls.

math.ST↗

arXiv2015

Bayesian nonparametric cross-study validation of prediction methods

We consider comparisons of statistical learning algorithms using multiple data sets, via leave-one-in cross-study validation: each of the algorithms is trained on one data set; the resulting model is then validated on each remaining data set. This poses two statistical challenges that need to be addressed simultaneously. The first is the assessment of study heterogeneity, with the aim of identifying a subset of studies within which algorithm comparisons can be reliably carried out. The second is the comparison of algorithms using the ensemble of data sets. We address both problems by integrating clustering and model comparison. We formulate a Bayesian model for the array of cross-study validation statistics, which defines clusters of studies with similar properties and provides the basis for meaningful algorithm comparison in the presence of study heterogeneity. We illustrate our approach through simulations involving studies with varying severity of systematic errors, and in the context of medical prognosis for patients diagnosed with cancer, using high-throughput measurements of the transcriptional activity of the tumor's genes.

stat.AP↗

arXiv2013

Hierarchical Bayesian analysis of somatic mutation data in cancer

Identifying genes underlying cancer development is critical to cancer biology and has important implications across prevention, diagnosis and treatment. Cancer sequencing studies aim at discovering genes with high frequencies of somatic mutations in specific types of cancer, as these genes are potential driving factors (drivers) for cancer development. We introduce a hierarchical Bayesian methodology to estimate gene-specific mutation rates and driver probabilities from somatic mutation data and to shed light on the overall proportion of drivers among sequenced genes. Our methodology applies to different experimental designs used in practice, including one-stage, two-stage and candidate gene designs. Also, sample sizes are typically small relative to the rarity of individual mutations. Via a shrinkage method borrowing strength from the whole genome in assessing individual genes, we reinforce inference and address the selection effects induced by multistage designs. Our simulation studies show that the posterior driver probabilities provide a nearly unbiased false discovery rate estimate. We apply our methods to pancreatic and breast cancer data, contrast our results to previous estimates and provide estimated proportions of drivers for these two types of cancer.

stat.AP↗

arXiv2012

Modeling dependent gene expression

In this paper we propose a Bayesian approach for inference about dependence of high throughput gene expression. Our goals are to use prior knowledge about pathways to anchor inference about dependence among genes; to account for this dependence while making inferences about differences in mean expression across phenotypes; and to explore differences in the dependence itself across phenotypes. Useful features of the proposed approach are a model-based parsimonious representation of expression as an ordinal outcome, a novel and flexible representation of prior information on the nature of dependencies, and the use of a coherent probability model over both the structure and strength of the dependencies of interest. We evaluate our approach through simulations and in the analysis of data on expression of genes in the Complement and Coagulation Cascade pathway in ovarian cancer.

stat.AP↗

arXiv2011

False discovery rates in somatic mutation studies of cancer

The purpose of cancer genome sequencing studies is to determine the nature and types of alterations present in a typical cancer and to discover genes mutated at high frequencies. In this article we discuss statistical methods for the analysis of somatic mutation frequency data generated in these studies. We place special emphasis on a two-stage study design introduced by Sjöblom et al. [Science 314 (2006) 268--274]. In this context, we describe and compare statistical methods for constructing scores that can be used to prioritize candidate genes for further investigation and to assess the statistical significance of the candidates thus identified. Controversy has surrounded the reliability of the false discovery rates estimates provided by the approximations used in early cancer genome studies. To address these, we develop a semiparametric Bayesian model that provides an accurate fit to the data. We use this model to generate a large collection of realistic scenarios, and evaluate alternative approaches on this collection. Our assessment is impartial in that the model used for generating data is not used by any of the approaches compared. And is objective, in that the scenarios are generated by a model that fits data. Our results quantify the conservative control of the false discovery rate with the Benjamini and Hockberg method compared to the empirical Bayes approach and the multiple testing method proposed in Storey [J. R. Stat. Soc. Ser. B Stat. Methodol. 64 (2002) 479--498]. Simulation results also show a negligible departure from the target false discovery rate for the methodology used in Sjöblom et al. [Science 314 (2006) 268--274].

stat.AP↗

arXiv2009

Statistical methods for automated drug susceptibility testing: Bayesian minimum inhibitory concentration prediction from growth curves

Determination of the minimum inhibitory concentration (MIC) of a drug that prevents microbial growth is an important step for managing patients with infections. In this paper we present a novel probabilistic approach that accurately estimates MICs based on a panel of multiple curves reflecting features of bacterial growth. We develop a probabilistic model for determining whether a given dilution of an antimicrobial agent is the MIC given features of the growth curves over time. Because of the potentially large collection of features, we utilize Bayesian model selection to narrow the collection of predictors to the most important variables. In addition to point estimates of MICs, we are able to provide posterior probabilities that each dilution is the MIC based on the observed growth curves. The methods are easily automated and have been incorporated into the Becton--Dickinson PHOENIX automated susceptibility system that rapidly and accurately classifies the resistance of a large number of microorganisms in clinical samples. Over seventy-five studies to date have shown this new method provides improved estimation of MICs over existing approaches.

stat.AP↗

arXiv2008

Hidden Markov models for the assessment of chromosomal alterations using high-throughput SNP arrays

Chromosomal DNA is characterized by variation between individuals at the level of entire chromosomes (e.g., aneuploidy in which the chromosome copy number is altered), segmental changes (including insertions, deletions, inversions, and translocations), and changes to small genomic regions (including single nucleotide polymorphisms). A variety of alterations that occur in chromosomal DNA, many of which can be detected using high density single nucleotide polymorphism (SNP) microarrays, are linked to normal variation as well as disease and are therefore of particular interest. These include changes in copy number (deletions and duplications) and genotype (e.g., the occurrence of regions of homozygosity). Hidden Markov models (HMM) are particularly useful for detecting such alterations, modeling the spatial dependence between neighboring SNPs. Here, we improve previous approaches that utilize HMM frameworks for inference in high throughput SNP arrays by integrating copy number, genotype calls, and the corresponding measures of uncertainty when available. Using simulated and experimental data, we, in particular, demonstrate how confidence scores control smoothing in a probabilistic framework. Software for fitting HMMs to SNP array data is available in the R package VanillaICE.

stat.AP↗

arXiv2008

Gamma shape mixtures for heavy-tailed distributions

An important question in health services research is the estimation of the proportion of medical expenditures that exceed a given threshold. Typically, medical expenditures present highly skewed, heavy tailed distributions, for which (a) simple variable transformations are insufficient to achieve a tractable low-dimensional parametric form and (b) nonparametric methods are not efficient in estimating exceedance probabilities for large thresholds. Motivated by this context, in this paper we propose a general Bayesian approach for the estimation of tail probabilities of heavy-tailed distributions, based on a mixture of gamma distributions in which the mixing occurs over the shape parameter. This family provides a flexible and novel approach for modeling heavy-tailed distributions, it is computationally efficient, and it only requires to specify a prior distribution for a single parameter. By carrying out simulation studies, we compare our approach with commonly used methods, such as the log-normal model and nonparametric alternatives. We found that the mixture-gamma model significantly improves predictive performance in estimating tail probabilities, compared to these alternatives. We also applied our method to the Medical Current Beneficiary Survey (MCBS), for which we estimate the probability of exceeding a given hospitalization cost for smoking attributable diseases. We have implemented the method in the open source GSM package, available from the Comprehensive R Archive Network.

stat.AP↗