Biostatistics

Upcoming Events

  • FDAWG Meeting: A two-stage classification for dealing with unseen...

    Tuesday, April 29, 2025
    4:00 PM to 5:00 PM
    Allan Rosenfield Building, 722 W. 168 St., New York, NY 10032 Room 627
    • Mailman School of Public Health
    On Tuesday, April 29th, from 4-5pm at the Functional Data Analysis Working Group (FDAWG) meeting, Dr. Ofer Harel from the University of Connecticut will give a talk titled “A two-stage classification for dealing with unseen clusters in the testing data"  Ofer Harel, Ph.D Dean of the College of Liberal Arts and Sciences, Professor in the Department of Statistics University of Connecticut A two-stage classification for dealing with unseen clusters in the testing data Abstract:  Classification is an important statistical tool that has increased its importance since the emergence of the data science revolution. However, a training data set that does not capture all underlying population subgroups (or clusters) will result in biased estimates or misclassification. In this presentation, we introduce a statistical and computational solution to a possible bias in classification when implemented on estimated population clusters. An unseen-cluster problem denotes the case in which the training data does not contain all underlying clusters in the population. Such a scenario may occur due to various reasons, such as sampling errors, selection bias, or emerging and disappearing population clusters. Once an unseen-cluster problem occurs, a testing observation will be misclassified because a classification rule based on the sample cannot capture a cluster not observed in the training data (sample). To overcome such issues, we suggest a two-stage classification method to ameliorate the unseen-cluster problem in classification. We suggest a test to identify the unseen-cluster problem and demonstrate the performance of the two-stage tailored classifier using simulations and a public data example. (This is a joint work with Jung Wun Lee).  

  • Biostats Lecture: Predictions of Advanced Adenoma...

    Thursday, May 1, 2025
    11:45 AM to 1:00 PM
    Allan Rosenfield Building, 722 W. 168 St., New York, NY 10032 Hess Commons
    • Mailman School of Public Health
    Thursday, May 1st 11:45am - 1:00pm, ARB Hess Commons Qing Pan, PhD Professor, Department of Biostatistics and Bioinformatics George Washington University, Milken Institute School of Public Health Predictions of Advanced Adenoma and High-Risk Pregnancies in Longitudinal Screening Studies Panel count data is common in cancer screening. In the context of colorectal cancer screening, our work focuses on the prediction of the probability of advanced adenoma conditional on patient-level risk factors and/or event history. We implement the joint frailty model proposed by Huang et al. (2006), which involves a non‐stationary Poisson process for recurrent adenoma events and informative screening time using semi‐parametric Cox models correlated by a latent frailty variable. Coefficients and baseline intensity functions are estimated through estimating equations. The subject-specific frailty value is estimated by the borrow‐strength method (Huang and Wang 2004). In addition, marginal models for the adenoma and screening events are also applicable when average covariate effects on the population level are of interest. Predictions of individual risks based on the marginal model and predictions based on the frailty models for patients with or without screening history are compared. When a patient’s screening history is available and sufficient adenoma events are observed, the predictions based on the frailty model with estimated subject‐specific frailty are superior. However, in the cases of early censoring when adenoma events are not observed for most patients or screening history is not available, the prediction based on the marginal model has better performance. For future screening, the individualized screening intervals based on the dynamic predictions of advanced adenoma risks will detect adenomas earlier with shorter lag times between adenoma occurrences compared to the current practice of fixed screening intervals for all. In a separate project, machine learning and deep learning models to identify pregnancies with elevated risks of adverse outcomes are compared. A novel GRU model that accommodates both static and time-varying information and allows interactions between these two kinds of covariates through additional attention layers provides better performance. Contributions of various types of covariates (questionnaires, blood tests, and ultrasound) to the prediction accuracy are compared for clinical practice in low- and middle-income countries.    Biostatistics Departmental Seminars & Lectures Lectures are in-person only unless marked otherwise. For all Zoom inquiries to attend seminars with a listed virtual option, please send an email to Erin Elliott, Programs Coordinator (ee2548@cumc.columbia.edu). During the Fall and Spring semesters, the Department of Biostatistics holds regular seminars on Thursdays, called the Levin Lecture Series, on a wide variety of topics which are of interest to both students and faculty. Over each semester, there are also often guest lectures outside the regular Thursday Levin Lecture Series, to provide a robust schedule the covers the wide range of topics in Biostatistics. The speakers are invited guests who spend the day of their seminar discussing their research with Biostatistics faculty and students. 

  • Biostatistics Practicum/APEx Symposium

    Friday, May 2, 2025
    12:00 PM to 5:00 PM
    Hammer Health Sciences Center, 701 W. 168 St., New York, NY 10032 Room 401
    • Mailman School of Public Health
    The Biostatistics MS and MPH Programs require students to complete one semester of hands-on, real-world experience through a Practicum (for MS students) or Applied Practice Experience, or APEx (for MPH students) This requirement gives students the chance to apply their classroom knowledge in a real-world public health setting—translating theory into practice. Each project is tailored to the student’s interests, skills, and career goals, and reflects a partnership with a public health organization or research initiative. It is a culmination of their academic journey at Columbia Mailman School of Public Health.  The Practicum/APEx Symposium is a showcase of this work. On Friday, May 2nd, students will present their projects to faculty, peers, and the broader Columbia community. It’s a celebration of their achievements and a day of learning that spans the breadth and depth of Biostatistics in action. We invite all CUIMC and Mailman School students, faculty, staff, and affiliates to join us in supporting and learning from the graduating Class of 2025. You can view the full titles and abstracts for each session organized by topic and room number on the event webpage, Topic Sections Include: Environmental Health Research, Data Analysis, Clinical Trials, Cancer Research, Deep Learning, Longitudinal Data Analysis, Missing Data , Topics in Mental Health, Drug Development, Data Visualizations, Data Analysis, Epidemiology, Machine Learning, Statistical Genetics, Bayesian Statistics, Causal Inference, Obersvational Study, and more... Schedule 12:00pm - 1:00pm: Welcome, Introduction, & Keynote Speaker (Hammer Room 401) (Sessions will be in Hammer Rooms 303, 305 & Lower Level 103, 106, 107, 108A, 109A, 202, 204, 205, 207, 210)  1:00pm - 2:00pm: Session 1  2:00pm - 2:30pm: Break 2:30pm - 3:30pm: Session 2 3:30pm - 3:45pm: Break 3:45pm - 4:45pm: Session 3 Keynote Speaker Hartaig Singh (Class of 2018) "Paths & Perspectives" Hartaig is currently a Staff Data Scientist at Zocdoc, where he leverages statistics and machine learning to drive strategic marketing insights. Since earning his MS in Biostatistics Theory & Methods from Columbia University’s Mailman School of Public Health in 2018, he has built a diverse career spanning consulting, healthcare, pharmaceuticals, and tech startups. His work has consistently focused on data science with a specialization in marketing analytics and enjoys combining rigorous statistical methods with real-world business impact.    

  • Biostats Lecture: Meta-clustering of Gene Expression Data

    Thursday, May 8, 2025
    11:45 AM to 1:00 PM
    Allan Rosenfield Building, 722 W. 168 St., New York, NY 10032 8th Floor Auditorium
    • Mailman School of Public Health
    Thursday, May 8th 11:45am - 1:00pm, ARB 8th Floor Auditorium Yingying Wei, PhD Associate Professor, Department of Statistics The Chinese University of Hong Kong Meta-clustering of Gene Expression Data Traditional meta-analyses pool effect sizes across studies to improve statistical power. Likewise, there is growing interest in joint clustering across datasets to identify disease subtypes for bulk gene expression data and to discover cell types for single-cell RNA-sequencing (scRNA-seq) data. Unfortunately, due to the prevalence of technical batch effects, directly clustering of samples from multiple gene expression datasets can lead to wrong results. Therefore, in the past several years, there has been very active research on the integration of multiple gene expression datasets. However, the discussion on when multiple gene expression datasets can be integrated for joint clustering is lacking. Obviously, if different subtypes are assayed in distinct batches, then meta-clustering would be impossible no matter what types of machine learning or statistical methods are used. In this talk, I will present our Batch-effects-correction-with-Unknown-Subtypes (BUS) framework. BUS is capable of adjusting batch effects explicitly, grouping samples that share similar characteristics into subtypes, identifying genes that distinguish subtypes and enjoying a linear-order computational complexity. The BUS framework can be adapted to perform meta-clustering for bulk gene expression data, scRNA-seq data collected from a single biological condition, and scRNA-seq data collected from multiple biological conditions, respectively. The proofs for model identifiability for the corresponding models provide insights on when multiple gene expression data can be integrated for meta-clustering and guidelines on experimental designs. Simulation studies and real data analyses show the advantages of our proposed models over state-of-the-art methods, especially when performing differential inference for scRNA-seq data collected from multiple conditions. Biostatistics Departmental Seminars & Lectures Lectures are in-person only unless marked otherwise. For all Zoom inquiries to attend seminars with a listed virtual option, please send an email to Erin Elliott, Programs Coordinator (ee2548@cumc.columbia.edu). During the Fall and Spring semesters, the Department of Biostatistics holds regular seminars on Thursdays, called the Levin Lecture Series, on a wide variety of topics which are of interest to both students and faculty. Over each semester, there are also often guest lectures outside the regular Thursday Levin Lecture Series, to provide a robust schedule the covers the wide range of topics in Biostatistics. The speakers are invited guests who spend the day of their seminar discussing their research with Biostatistics faculty and students.