Avatar

Shariq Mohammed

Assistant Professor

Department of Biostatistics

Boston University

Biography

I am an Assistant Professor in the Department of Biostatistics at Boston University School of Public Health. Prior to joining BU, I was a postdoctoral research fellow in the Departments of Biostatistics, and Computational Medicine and Bioinformatics, and a Precision Health Scholar at The University of Michigan-Ann Arbor. I obtained my PhD in Statistics from University of Connecticut in 2018.

My research interests include Bayesian modeling, variable selection, geometric functional data analysis and applications to biomedical imaging data. My current research is focused on building statistical methods to address relevant questions in different disease contexts, by integrating data from multiple platforms.

Interests

  • Bayesian Modeling
  • Variable Selection
  • Medical Imaging Analysis
  • Cancer Imaging Genomics
  • Precision Health

Education

  • Ph.D. in Statistics, 2018

    University of Connecticut

  • M.S. in Statistics, 2017

    University of Connecticut

  • M.Sc. in Applications of Mathematics, 2014

    Chennai Mathematical Institute

  • B.Math.(Hons.), 2012

    Indian Statistical Institute

Research

.js-id-ImgGen

Geometric functional data analysis in biomedical imaging

We are working on approaches to build classification algorithms for imaging data from Ophthalmology to assess retinal changes in diabetic retinopathy. We are also working on leveraging the geometric functional data analysis approaches in conjunction with spatial analysis techniques for other medical imaging data such as radiological and histo-pathological imaging. The applications include imaging data from gliomas, pancreatic cancer and immunotherapy for lung cancer.

Radiogenomic analysis in gliomas

We are building Bayesian variable selection approaches to identify associations between molecular characteristics and imaging heterogeneity. These assocaitions are also studied while considering (by mimicking) the tumor growth process. We are specifically interested in radiological-imaging and the cancer driver-genes/pathways of lower grade gliomas. We are employing methods from geometric functional data analysis. We are also working on methods for compositional data to assess associations of genomics and the volumetric characteristics of the tumor.

Spatial modeling of large insurance claims and occurrence data

We are working on building efficient spatial risk-quantification models in insurance claims data. We are also building graphical Tweedie compound Poisson double generalized linear models.

Publications

.js-id-Selected-Journal-Articles

Tumor radiogenomics with Bayesian layered variable selection

Abstract: We propose a statistical framework to integrate radiological magnetic resonance imaging (MRI) and genomic data to identify …

Spatial Tweedie exponential dispersion models: An application to insurance rate-making

Scandinavian Actuarial Journal (2021). (Just Accepted) Abstract: In this paper we propose a statistical modeling framework that …

RADIOHEAD: Radiogenomic analysis incorporating tumor heterogeneity in imagine through densities

Annals of Applied Statistics (2021). (Just Accepted) Abstract: Recent technological advancements have enabled detailed investigation of …

Network-based Modeling of COVID-19 Dynamics: Early Pandemic Spread in India

Abstract: Modeling the dynamics of COVID-19 pandemic spread is a challenging and relevant problem. Established models for the epidemic …

Scalable spatio-temporal Bayesian analysis of high-dimensional electroencephalography data

Canadian Journal of Statistics (2021). Abstract: We present a scalable Bayesian modeling approach for identifying brain regions that …

A Bayesian 2D functional linear model for gray-level co-occurrence matrices in texture analysis of lower grade gliomas

NeuroImage: Clinical (2020). Abstract: In cancer radiomics, textural features evaluated from image intensity-derived gray-level …

Classification of high-dimensional electroencephalography data with location selection using structured spike-and-slab prior

Statistical Analysis and Data Mining: The ASA Data Science Journal (2020). Abstract: With the advent of modern technologies, it is …

Integrative Bayesian models using post-selective inference: a case study in radiogenomics

Abstract: Identifying direct links between gene pathways and clinical endpoints for highly fatal diseases such as cancer is a …

Biomedical applications of geometric functional data analysis

Handbook of Variational Methods for Nonlinear Geometric Data - Springer (2020). Abstract: In this chapter, we describe several …

Spatial risk estimation in Tweedie compound Poisson double generalized linear models

Abstract: Tweedie exponential dispersion family constitutes a fairly rich sub-class of the celebrated exponential family. In …

Bayesian variable selection using spike-and-slab priors with application to high dimensional electroencephalography data by local modelling

Journal of the Royal Statistical Society: Series C (Applied Statistics) (2019). Abstract: Because of the immense technological …

Assessing malaria using neutral zone classifiers with mixture discriminant analysis on 2D images of red blood cells

Journal of Biostatistics and Epidemiology (2019). Abstract: Background and Aim: We aim to build a classifier to distinguish between …

A dynamical systems approach to systemic risk in a financial network

In 2016 Indian Control Conference (ICC). Abstract: The insolvency of a financial entity such as a bank can trigger a sequence of …

Contact

  • shariqm at bu dot edu
  • Crosstown Center (CT-303), 801 Massachusetts Ave, Boston, MA 02118