Featured News

Vaccine and Infectious Disease Division

Featured News

Laboratories inside the Vaccine and Infectious Disease Division (VIDD), Fred Hutchinson Cancer Research Center

COMPASS paper wins Mitchell Prize

The International Society for Bayesian Analysis (ISBA) sponsors the Mitchell Prize, which is awarded in recognition of an outstanding paper that describes how a Bayesian analysis has solved an important applied problem. The 2016 Mitchell Prize, announced at the Joint Statistical Meetings on August 7, 2017, went to the paper “COMPASS Identifies T-cell Subsets Correlated with Clinical Outcomes” published in Nature Biotechnology. This publication, coming from Dr. Raphael Gottardo’s group in VIDD, describes combinatorial polyfunctionality analysis of antigen-specific T-cell subsets (COMPASS), a novel computational system for deciphering immune responses that has shed new light on correlates of protection (COP) from the RV144 HV vaccine trial.

The power to interrogate cells on the single cell level is becoming more and more apparent as these studies are providing greater insight into biological processes such as immune responses to vaccination. In any biological sample, cells are not homogeneous; even differences in cell cycle can distort gene expression analyses when analyzed ‘in bulk.’ With the arrival of next generation single cell and flow cytometry/intracellular cytokine staining (ICS) technologies comes a great need for tools to analyze the surfeit of data that is being generated. This is especially germane to data generated from large scale randomized clinical trials. 

Raphael Gottardo

Dr. Raphael Gottardo, member of VIDD, led the study.

This paper by Lin et al. has uncovered and displayed the importance of polyfunctional immune responses to HIV vaccination. The wealth of information stemming from follow-up studies on RV144 motivated scientists to develop better statistical and high throughput tools for interpreting the data. The sheer volume of data, the complexity of it and need to ‘catch’ antigen responses that may be at very low numbers within a patient’s total immune response are all factors. In the total T-cell population from PBMC isolated from any trial subject only a small fraction will be HIV specific. Within this rare group are T cells that target different viral epitopes, express different cytokine combinations, and have different immune function. Therefore, the numbers become very small and thus a need to find the proverbial needle in the haystack is crucial to determine important COP, not just for HIV but any important clinical endpoint. COMPASS can be used to compute the functionality score for each individual, defined as the proportion of vaccine-mediated antigen-specific T-cell subsets detected among all possible cytokine producing subsets. This statistical tool has uncovered differences in T-cell polyfunctionality in response to multiple antigens in TB and HIV disease, the latter of which has now been determined as a separate COP. Importantly, COMPASS provides meaning to complex multidimensional immunological readouts that allow one to actually now define polyfunctionality. This is critical as all T cells and all cytokines are not created equal. This tool to allow both rational description and analytical evaluation of antigen specific T cells is a major breakthrough for the field.

This Bayesian analysis methodology has already been applied to multiple other studies, highlighting the impact has and will have on clinical trial COP analysis, which is necessary for designing effective immunotherapies and vaccines.

VIDD scientists who co-authored the paper are Lin Lin, Greg Finak, Kevin Ushey, Nicole Frahm, Peter Gilbert, Stephen De Rosa, Julie McElrath, and Raphael Gottardo.


Lin L, Finak G, Ushey K, Seshadri C, Hawn TR, Frahm N, Scriba TJ, Mahomed H, Hanekom W, Bart PA, Pantaleo G, Tomaras GD, Rerks-Ngarm S, Kaewkungwal J, Nitayaphan S, Pitisuttithum P, Michael NL, Kim JH, Robb ML, O'Connell RJ, Karasavvas N, Gilbert P, C De Rosa S, McElrath MJ, Gottardo R. COMPASS identifies T-cell subsets correlated with clinical outcomes. Nat Biotechnol. 2015 Jun;33(6):610-6.