CalR: A toolkit and repository for experiments of energy homeostasis using indirect calorimetry

Project Summary
Modern indirect calorimetry systems allow for high-density multi-dimensional time-series measurements of
components affecting body weight and energy homeostasis. Indirect calorimetry is used for understanding the
factors influencing pathogenic changes to body weight—examples include weight loss with cancer cachexia, or
weight gain with obesity. These indirect calorimetry systems generate a flood of raw data that requires time-
consuming manual manipulation for formatting, data cleaning, quality control, and visualization. Beyond data
handling, analysis of indirect calorimetry experiments requires specialized statistical treatment to account for
differential contributions of fat mass and lean mass to metabolic rates. Surprisingly, no tools or resources exist
to address these shortcomings. Comparisons between experiments are rarely performed due to the different
types of instruments with varying units of measurement and ad hoc statistical treatments of data. To address
this critical need, we propose the creation of a free online tool, CalR, that helps scientists quickly and efficiently
analyze indirect calorimetry data by providing standardized methods for reproducible research and a site to
store and aggregate datasets. The preliminary version of CalR we launched is a user-friendly but sophisticated
web tool that uses a graphical user interface to import data files from different instruments, quickly visualize
experimental results, and perform basic statistical analyses. After several years of iterative development, in
addition to constructive feedback from a user survey, it is clear that additional functionality and statistical
methods need to be developed to realize the potential of this project. The broad goal of this research is the
development of a framework that delivers modern tools for the analysis of the physiological data affecting body
weight. These new tools, built on the existing CalR software, will continue to be freely provided to the scientific
community. Specific aims will establish three aspects of this framework: 1) New analysis features intended to
improve prediction of body weight change, automatic determination of metabolic flexibility, improved
determination of metabolic rates, and better visualizations. 2) New statistical methods incorporating additional
covariates into analysis and a module to determine statistical power for metabolic experiments. 3) The
development of a repository where indirect calorimetry data can be deposited and made broadly accessible for
large-scale analysis. For each aim, we will develop and thoroughly test each component, after which we will
provide a functional web tool online to share with the thousands of scientists who use our free software
platform. These efforts will extract more nuanced information and will improve the interpretation of each
experiment for users in all fields where body weight impacts disease processes.

Grant Number: 5R01DK133948-04
NIH Institute/Center: NIH
Principal Investigator: ALEXANDER BANKS