Phenotyping & Visualization

DPdash is a web application that allows users to visualize their data on a desktop, laptop, or mobile device. It reads in data saved to a local filesystem, and constructs a matrix for each subject based on user configurations. Visualizing these matrices allow investigators to verify that data are being captured successfully and explore relationships across data streams.

Main Page

The main page provides a list of studies and their subjects. Each item in the study list contains the name and number of days of the study, and the number of subjects currently enrolled in the study. Subjects can be sorted based on the number of days, updated dates, and subject ID. Users can filter the list of subjects with the menu on the right, and highlight subjects of main focus by clicking a star button next to their subject ID.

Dashboard

Each cell of the matrix component is colored based on the value range of the variable. A white cell indicates a missing data point, which could reflect a data capturing issue or a pipeline execution issue. The number displayed inside a cell, if present, shows the actual value of the variable. Each dashboard comes with a statistics table with the minimum, maximum, mean, and the percentage of missing values for each variable.

Personalization

Users have the ability to flexibly configure their dashboard to focus on their specific research goals or phenotypes of interest. The mapping between colors and the numeric range of each variable determines the view the user will see when daily data are visualized on DPdash.

Security

DPdash provides secure log-in and registration pages to ensure user data are protected. Admins can monitor user activity and manage project access permissions according to IRB policy.

The primary goal of DPdash is to resolve common challenges that researchers face in cross-sectional or longitudinal studies that require assessment of high-dimensional and unstructured data. DPdash streamlines the process of quality control with user-driven web visualizations that highlight variations in phenotypes.