Debashis Ghosh
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Debashis Ghosh has a total of 40 co-authors. Here are the top co-authors by number of shared publications:
- Hyungwon Choi
Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore, Singapore
5 shared publications - Zhaohui Steve Qin
Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University, Atlanta, Georgia, USA
4 shared publications - Arul M. Chinnaiyan
Michigan Center for Translational Pathology, University of Michigan, Ann Arbor, MI, USA.
3 shared publications Jeremy M. G. Taylor
Department of Biostatistics, University of Michigan, Ann Arbor, Michigan 48103, U.S.A.
3 shared publications- Yu Zhang
Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Department of Thermal Energy Engineering, Tsinghua University, Beijing 100084, China
3 shared publications - Alexey Nesvizhskii
Department of Pathology, University of Michigan, Ann Arbor, MI, USA
2 shared publications
Latest Publications
- 29 August 2014
- Published by Walter de Gruyter GmbH in Journal of Causal Inference
- Vol. 3 (1) , 25-40
A Boosting Algorithm for Estimating Generalized Propensity Scores with Continuous Treatments
In this article, we study the causal inference problem with a continuous treatment variable using propensity score-based methods. For a continuous treatment, the generalized propensity score is defined as the conditional density of the treatment-level given covariates (confounders). The dose–response function is then estimated by inverse probability weighting, where the weights are calculated from the estimated propensity scores. When the dimension of the covariates is large, the traditional nonparametric density estimation suffers from the curse of dimensionality. Some researchers have suggested a two-step estimation procedure by first modeling the mean function. In this study, we suggest a boosting algorithm to estimate the mean function of the treatment given covariates. In boosting, an important tuning parameter is the number of trees to be generated, which essentially determines the trade-off between bias and variance of the causal estimator. We propose a criterion called average absolute correlation coefficient (AACC) to determine the optimal number of trees. Simulation results show that the proposed approach performs better than a simple linear approximation or L2 boosting. The proposed methodology is also illustrated through the Early Dieting in Girls study, which examines the influence of mothers’ overall weight concern on daughters’ dieting behavior.
- 29 August 2014
- Published by Walter de Gruyter GmbH in Journal of Causal Inference
- Vol. 3 (1) , 25-40
A Boosting Algorithm for Estimating Generalized Propensity Scores with Continuous Treatments
In this article, we study the causal inference problem with a continuous treatment variable using propensity score-based methods. For a continuous treatment, the generalized propensity score is defined as the conditional density of the treatment-level given covariates (confounders). The dose–response function is then estimated by inverse probability weighting, where the weights are calculated from the estimated propensity scores. When the dimension of the covariates is large, the traditional nonparametric density estimation suffers from the curse of dimensionality. Some researchers have suggested a two-step estimation procedure by first modeling the mean function. In this study, we suggest a boosting algorithm to estimate the mean function of the treatment given covariates. In boosting, an important tuning parameter is the number of trees to be generated, which essentially determines the trade-off between bias and variance of the causal estimator. We propose a criterion called average absolute correlation coefficient (AACC) to determine the optimal number of trees. Simulation results show that the proposed approach performs better than a simple linear approximation or L2 boosting. The proposed methodology is also illustrated through the Early Dieting in Girls study, which examines the influence of mothers’ overall weight concern on daughters’ dieting behavior.
- 26 August 2014
- Published by Elsevier in Statistical Methodology
- Vol. 22, 47-57
Incorporating auxiliary information for improved prediction using combination of kernel machines
- 26 August 2014
- Published by Elsevier in Statistical Methodology
- Vol. 22, 47-57
Incorporating auxiliary information for improved prediction using combination of kernel machines
- Open Access
- 28 June 2014
- Published by Springer Nature in BMC Bioinformatics
- Vol. 15 (1) , 1-12
Meta-analysis based on weighted ordered P-values for genomic data with heterogeneity
Meta-analysis has become increasingly popular in recent years, especially in genomic data analysis, due to the fast growth of available data and studies that target the same questions. Many methods have been developed, including classical ones such as Fisher’s combined probability test and Stouffer’s Z-test. However, not all meta-analyses have the same goal in mind. Some aim at combining information to find signals in at least one of the studies, while others hope to find more consistent signals across the studies. While many classical meta-analysis methods are developed with the former goal in mind, the latter goal has much more practicality for genomic data analysis. In this paper, we propose a class of meta-analysis methods based on summaries of weighted ordered p-values (WOP) that aim at detecting significance in a majority of studies. We consider weighted versions of classical procedures such as Fisher’s method and Stouffer’s method where the weight for each p-value is based on its order among the studies. In particular, we consider weights based on the binomial distribution, where the median of the p-values are weighted highest and the outlying p-values are down-weighted. We investigate the properties of our methods and demonstrate their strengths through simulations studies, comparing to existing procedures. In addition, we illustrate application of the proposed methodology by several meta-analysis of gene expression data. Our proposed weighted ordered p-value (WOP) methods displayed better performance compared to existing methods for testing the hypothesis that there is signal in the majority of studies. They also appeared to be much more robust in applications compared to the rth ordered p-value (rOP) method (Song and Tseng, Ann. Appl. Stat. 2014, 8(2):777–800). With the flexibility of incorporating different p-value combination methods and different weighting schemes, the weighted ordered p-values (WOP) methods have great potential in detecting consistent signal in meta-analysis with heterogeneity.
- 28 June 2014
- Published by Springer Nature in BMC Bioinformatics
- Vol. 15 (1) , 1-12
Open AccessMeta-analysis based on weighted ordered P-values for genomic data with heterogeneity
Meta-analysis has become increasingly popular in recent years, especially in genomic data analysis, due to the fast growth of available data and studies that target the same questions. Many methods have been developed, including classical ones such as Fisher’s combined probability test and Stouffer’s Z-test. However, not all meta-analyses have the same goal in mind. Some aim at combining information to find signals in at least one of the studies, while others hope to find more consistent signals across the studies. While many classical meta-analysis methods are developed with the former goal in mind, the latter goal has much more practicality for genomic data analysis. In this paper, we propose a class of meta-analysis methods based on summaries of weighted ordered p-values (WOP) that aim at detecting significance in a majority of studies. We consider weighted versions of classical procedures such as Fisher’s method and Stouffer’s method where the weight for each p-value is based on its order among the studies. In particular, we consider weights based on the binomial distribution, where the median of the p-values are weighted highest and the outlying p-values are down-weighted. We investigate the properties of our methods and demonstrate their strengths through simulations studies, comparing to existing procedures. In addition, we illustrate application of the proposed methodology by several meta-analysis of gene expression data. Our proposed weighted ordered p-value (WOP) methods displayed better performance compared to existing methods for testing the hypothesis that there is signal in the majority of studies. They also appeared to be much more robust in applications compared to the rth ordered p-value (rOP) method (Song and Tseng, Ann. Appl. Stat. 2014, 8(2):777–800). With the flexibility of incorporating different p-value combination methods and different weighting schemes, the weighted ordered p-values (WOP) methods have great potential in detecting consistent signal in meta-analysis with heterogeneity.
- Open Access
- 24 June 2014
- Published by Springer Nature in Statistics in Biosciences
- Vol. 7 (2) , 262-281
An Adaptive Genetic Association Test Using Double Kernel Machines
Recently, gene set-based approaches have become very popular in gene expression profiling studies for assessing how genetic variants are related to disease outcomes. Since most genes are not differentially expressed, existing pathway tests considering all genes within a pathway suffer from considerable noise and power loss. Moreover, for a differentially expressed pathway, it is of interest to select important genes that drive the effect of the pathway. In this article, we propose an adaptive association test using double kernel machines (DKM), which can both select important genes within the pathway as well as test for the overall genetic pathway effect. This DKM procedure first uses the garrote kernel machines test for the purposes of subset selection and then the least squares kernel machine test for testing the effect of the subset of genes. An appealing feature of the kernel machine framework is that it can provide a flexible and unified method for multi-dimensional modeling of the genetic pathway effect allowing for both parametric and nonparametric components. This DKM approach is illustrated with application to simulated data as well as to data from a neuroimaging genetics study.
- 24 June 2014
- Published by Springer Nature in Statistics in Biosciences
- Vol. 7 (2) , 262-281
Open AccessAn Adaptive Genetic Association Test Using Double Kernel Machines
Recently, gene set-based approaches have become very popular in gene expression profiling studies for assessing how genetic variants are related to disease outcomes. Since most genes are not differentially expressed, existing pathway tests considering all genes within a pathway suffer from considerable noise and power loss. Moreover, for a differentially expressed pathway, it is of interest to select important genes that drive the effect of the pathway. In this article, we propose an adaptive association test using double kernel machines (DKM), which can both select important genes within the pathway as well as test for the overall genetic pathway effect. This DKM procedure first uses the garrote kernel machines test for the purposes of subset selection and then the least squares kernel machine test for testing the effect of the subset of genes. An appealing feature of the kernel machine framework is that it can provide a flexible and unified method for multi-dimensional modeling of the genetic pathway effect allowing for both parametric and nonparametric components. This DKM approach is illustrated with application to simulated data as well as to data from a neuroimaging genetics study.
- 17 June 2014
- Published by Springer Nature
- p. 405-422
Statistical Considerations in the Analysis of Rare Variants
- 17 June 2014
- Published by Springer Nature
- p. 405-422
Statistical Considerations in the Analysis of Rare Variants
- Open Access
- 12 May 2014
- Published by Springer Nature in BMC Genomics
- Vol. 15 (1) , 1-15
Regression hidden Markov modeling reveals heterogeneous gene expression regulation: a case study in mouse embryonic stem cells
Studies have shown the strong association between histone modification levels and gene expression levels. The detailed relationships between the two can vary substantially due to differential regulation, and hence a simple regression model may not be adequate. We apply a regression hidden Markov model (regHMM) to further investigate the potential multiple relationships between genes and histone methylation levels in mouse embryonic stem cells. Seven histone methylation levels are used in the study. Averaged histone modifications over non-overlapping 200 bp windows on the range transcription starting site (TSS) ± 1 Kb are used as predictors, and in total 70 explanatory variables are generated. Based on regHMM results, genes segregated into two groups, referred to as State 1 and State 2, have distinct association strengths. Genes in State 1 are better explained by histone methylation levels with R2=.72 while those in State 2 have weaker association strength with R2=.38. The regression coefficients in the two states are not very different in magnitude except in the intercept,.25 and 1.15 for State 1 and State 2, respectively. We found specific GO categories that may be attributed to the different relationships. The GO categories more frequently observed in State 2 match those of housekeeping genes, such as cytoplasm, nucleus, and protein binding. In addition, the housekeeping gene expression levels are significantly less explained by histone methylation in mouse embryonic stem cells, which is consistent with the constitutive expression patterns that would be expected. Gene expression levels are not universally affected by histone methylation levels, and the relationships between the two differ by the gene functions. The expression levels of the genes that perform the most common housekeeping genes’ GO categories are less strongly associated with histone methylation levels. We suspect that additional biological factors may also be strongly associated with the gene expression levels in State 2. We discover that the effect of the presence of CpG island in TSS ± 1 Kb is larger in State 2.
- 12 May 2014
- Published by Springer Nature in BMC Genomics
- Vol. 15 (1) , 1-15
Open AccessRegression hidden Markov modeling reveals heterogeneous gene expression regulation: a case study in mouse embryonic stem cells
Studies have shown the strong association between histone modification levels and gene expression levels. The detailed relationships between the two can vary substantially due to differential regulation, and hence a simple regression model may not be adequate. We apply a regression hidden Markov model (regHMM) to further investigate the potential multiple relationships between genes and histone methylation levels in mouse embryonic stem cells. Seven histone methylation levels are used in the study. Averaged histone modifications over non-overlapping 200 bp windows on the range transcription starting site (TSS) ± 1 Kb are used as predictors, and in total 70 explanatory variables are generated. Based on regHMM results, genes segregated into two groups, referred to as State 1 and State 2, have distinct association strengths. Genes in State 1 are better explained by histone methylation levels with R2=.72 while those in State 2 have weaker association strength with R2=.38. The regression coefficients in the two states are not very different in magnitude except in the intercept,.25 and 1.15 for State 1 and State 2, respectively. We found specific GO categories that may be attributed to the different relationships. The GO categories more frequently observed in State 2 match those of housekeeping genes, such as cytoplasm, nucleus, and protein binding. In addition, the housekeeping gene expression levels are significantly less explained by histone methylation in mouse embryonic stem cells, which is consistent with the constitutive expression patterns that would be expected. Gene expression levels are not universally affected by histone methylation levels, and the relationships between the two differ by the gene functions. The expression levels of the genes that perform the most common housekeeping genes’ GO categories are less strongly associated with histone methylation levels. We suspect that additional biological factors may also be strongly associated with the gene expression levels in State 2. We discover that the effect of the presence of CpG island in TSS ± 1 Kb is larger in State 2.
- Open Access
- 14 April 2014
- Published by Springer Nature in BMC Bioinformatics
- Vol. 15 (1) , 108
A two-step hierarchical hypothesis set testing framework, with applications to gene expression data on ordered categories
In complex large-scale experiments, in addition to simultaneously considering a large number of features, multiple hypotheses are often being tested for each feature. This leads to a problem of multi-dimensional multiple testing. For example, in gene expression studies over ordered categories (such as time-course or dose-response experiments), interest is often in testing differential expression across several categories for each gene. In this paper, we consider a framework for testing multiple sets of hypothesis, which can be applied to a wide range of problems.
- 14 April 2014
- Published by Springer Nature in BMC Bioinformatics
- Vol. 15 (1) , 108
Open AccessA two-step hierarchical hypothesis set testing framework, with applications to gene expression data on ordered categories
In complex large-scale experiments, in addition to simultaneously considering a large number of features, multiple hypotheses are often being tested for each feature. This leads to a problem of multi-dimensional multiple testing. For example, in gene expression studies over ordered categories (such as time-course or dose-response experiments), interest is often in testing differential expression across several categories for each gene. In this paper, we consider a framework for testing multiple sets of hypothesis, which can be applied to a wide range of problems.
- Open Access
- 19 February 2014
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 30 (12) , 1755-1756
MRHMMs: Multivariate Regression Hidden Markov Models and the variantS
Summary: Hidden Markov models (HMMs) are flexible and widely used in scientific studies. Particularly in genomics and genetics, there are multiple distinct regimes in the genome within each of which the relationships among multivariate features are distinct. Examples include differential gene regulation depending on gene functions and experimental conditions, and varying combinatorial patterns of multiple transcription factors. We developed a software package called MRHMMs (Multivariate Regression Hidden Markov Models and the variantS) that accommodates a variety of HMMs that can be flexibly applied to many biological studies and beyond. MRHMMs supplements existing HMM software packages in two aspects. First, MRHMMs provides a diverse set of emission probability structures, including mixture of multivariate normal distributions and (logistic) regression models. Second, MRHMMs is computationally efficient for analyzing large data-sets generated in current genome-wide studies. Especially, the software is written in C for the speed advantage and further amenable to implement alternative models to meet users’ own purposes. Availability and implementation: http://sourceforge.net/projects/mrhmms/ Contact:ghoshd@psu.edu or yuzhang@stat.psu.edu Supplementary information: Supplementary data are available at Bioinformatics online.
- 19 February 2014
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 30 (12) , 1755-1756
Open AccessMRHMMs: Multivariate Regression Hidden Markov Models and the variantS
Summary: Hidden Markov models (HMMs) are flexible and widely used in scientific studies. Particularly in genomics and genetics, there are multiple distinct regimes in the genome within each of which the relationships among multivariate features are distinct. Examples include differential gene regulation depending on gene functions and experimental conditions, and varying combinatorial patterns of multiple transcription factors. We developed a software package called MRHMMs (Multivariate Regression Hidden Markov Models and the variantS) that accommodates a variety of HMMs that can be flexibly applied to many biological studies and beyond. MRHMMs supplements existing HMM software packages in two aspects. First, MRHMMs provides a diverse set of emission probability structures, including mixture of multivariate normal distributions and (logistic) regression models. Second, MRHMMs is computationally efficient for analyzing large data-sets generated in current genome-wide studies. Especially, the software is written in C for the speed advantage and further amenable to implement alternative models to meet users’ own purposes. Availability and implementation: http://sourceforge.net/projects/mrhmms/ Contact:ghoshd@psu.edu or yuzhang@stat.psu.edu Supplementary information: Supplementary data are available at Bioinformatics online.
- Open Access
- 16 January 2013
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 29 (5) , 533-541
Sparsely correlated hidden Markov models with application to genome-wide location studies
Motivation: Multiply correlated datasets have become increasingly common in genome-wide location analysis of regulatory proteins and epigenetic modifications. Their correlation can be directly incorporated into a statistical model to capture underlying biological interactions, but such modeling quickly becomes computationally intractable. Results: We present sparsely correlated hidden Markov models (scHMM), a novel method for performing simultaneous hidden Markov model (HMM) inference for multiple genomic datasets. In scHMM, a single HMM is assumed for each series, but the transition probability in each series depends on not only its own hidden states but also the hidden states of other related series. For each series, scHMM uses penalized regression to select a subset of the other data series and estimate their effects on the odds of each transition in the given series. Following this, hidden states are inferred using a standard forward–backward algorithm, with the transition probabilities adjusted by the model at each position, which helps retain the order of computation close to fitting independent HMMs (iHMM). Hence, scHMM is a collection of inter-dependent non-homogeneous HMMs, capable of giving a close approximation to a fully multivariate HMM fit. A simulation study shows that scHMM achieves comparable sensitivity to the multivariate HMM fit at a much lower computational cost. The method was demonstrated in the joint analysis of 39 histone modifications, CTCF and RNA polymerase II in human CD4+ T cells. scHMM reported fewer high-confidence regions than iHMM in this dataset, but scHMM could recover previously characterized histone modifications in relevant genomic regions better than iHMM. In addition, the resulting combinatorial patterns from scHMM could be better mapped to the 51 states reported by the multivariate HMM method of Ernst and Kellis. Availability: The scHMM package can be freely downloaded from http://sourceforge.net/p/schmm/ and is recommended for use in a linux environment. Contact: ghoshd@psu.edu or zhaohui.qin@emory.edu Supplementary information: Supplementary data are available at Bioinformatics online.
- 16 January 2013
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 29 (5) , 533-541
Open AccessSparsely correlated hidden Markov models with application to genome-wide location studies
Motivation: Multiply correlated datasets have become increasingly common in genome-wide location analysis of regulatory proteins and epigenetic modifications. Their correlation can be directly incorporated into a statistical model to capture underlying biological interactions, but such modeling quickly becomes computationally intractable. Results: We present sparsely correlated hidden Markov models (scHMM), a novel method for performing simultaneous hidden Markov model (HMM) inference for multiple genomic datasets. In scHMM, a single HMM is assumed for each series, but the transition probability in each series depends on not only its own hidden states but also the hidden states of other related series. For each series, scHMM uses penalized regression to select a subset of the other data series and estimate their effects on the odds of each transition in the given series. Following this, hidden states are inferred using a standard forward–backward algorithm, with the transition probabilities adjusted by the model at each position, which helps retain the order of computation close to fitting independent HMMs (iHMM). Hence, scHMM is a collection of inter-dependent non-homogeneous HMMs, capable of giving a close approximation to a fully multivariate HMM fit. A simulation study shows that scHMM achieves comparable sensitivity to the multivariate HMM fit at a much lower computational cost. The method was demonstrated in the joint analysis of 39 histone modifications, CTCF and RNA polymerase II in human CD4+ T cells. scHMM reported fewer high-confidence regions than iHMM in this dataset, but scHMM could recover previously characterized histone modifications in relevant genomic regions better than iHMM. In addition, the resulting combinatorial patterns from scHMM could be better mapped to the 51 states reported by the multivariate HMM method of Ernst and Kellis. Availability: The scHMM package can be freely downloaded from http://sourceforge.net/p/schmm/ and is recommended for use in a linux environment. Contact: ghoshd@psu.edu or zhaohui.qin@emory.edu Supplementary information: Supplementary data are available at Bioinformatics online.
- 3 January 2013
- Published by Springer Nature
- Vol. 972, 141-153
Genomic Outlier Detection in High-Throughput Data Analysis
In the analysis of high-throughput data, a very common goal is the detection of genes or of differential expression between two groups or classes. A recent finding from the scientific literature in prostate cancer demonstrates that by searching for a different...
- 3 January 2013
- Published by Springer Nature
- Vol. 972, 141-153
Genomic Outlier Detection in High-Throughput Data Analysis
In the analysis of high-throughput data, a very common goal is the detection of genes or of differential expression between two groups or classes. A recent finding from the scientific literature in prostate cancer demonstrates that by searching for a different...