With SEND Explorer Plus and Warehouse, study data is loaded to a central database for each client. This database holds all of the data included for each study, including any data collected on control animals.
Historical control data (HCD) is a valuable tool in preclinical data analysis. HCD is routinely aggregated to establish reference ranges for clinical pathology and other numeric endpoints measured in nonclinical studies. These reference ranges are generally lab specific and are used as a comparator for helping to determine the relevance of toxicological effects. HCD is also used for qualitative data such as histopathology, to help determine background incidence of findings. Another use for HCD is to provide a baseline for comparison to ensure the reliability and quality of a test system.
SEND Explorer has several visualizations that allow users to analyze and visualize historical control data contained in the SEND Explorer database, all available from the top Visualizations menu.
Historical Control Distribution Values Visualization
The multi-study Historical Control Values Distribution visualization in SEND Explorer is an excellent tool for developing historical control reference ranges for specified endpoints. This visualization calculates and graphs the distribution of values for a selected endpoint across studies for untreated control subjects. To select an endpoint, click on the Test Codes button on the top left. The range of observed values for the endpoint is divided into equally sized intervals and the number of control observations that fall within each interval is depicted in a distribution histogram. For a given distribution, the user can plot and/or color by any of a number of metadata values to highlight distribution differences based on metadata attributes.
A number of options are available to customize the visualization using the configuration options on the left-hand side of the screen, including altering the width or height of the plot, the number of equal intervals, the scaling for the y-axis, the display on mouse-over, and whether the color-by options produce stacked or adjacent bars. The user can set the x-axis min and max value themselves or use the data-driven min and max. The user can determine whether to include all available observations or just one per subject (which is defined as the last observation for that subject). The user can determine whether to include pretest measures for treated subjects in addition to measures from control animals. The user can filter the visualization to selected observations based on selected metadata values.
A summary table displays beneath the visualization, identifying summary measures associated with the distributions including n, mean, minimum, maximum, and standard deviation, plus several percentiles. All of the data contributing to the distribution plots for the selected endpoint can be exported to Excel/CSV or selected data from the distribution can be displayed in a grid beneath the visualization and subsequently exported to Excel/CSV.
Scatterplot Timecourse Historical Control Reference Overlay
The Scatterplot Timecourse visualization includes a Historical Control Reference option which allows the user to specify criteria for identifying an appropriate set of control subjects from which a relevant reference range can be established. If this option is enabled (click the checkbox next to Historical Control Reference under Display Options on the left), the resulting historical reference range, shaded in grey, will be overlayed on the Multi-Study Scatterplots, matched by species. The legend identifies the range of values relevant for each species. The reference range can be filtered for selected species to limit the species included in the legend by clicking on the gear icon to the right of the Historical Control Reference checkbox where an option appears to filter by Species (and many other fields). If Sex is included as a Plot By field, the historical control reference ranges in the legend will be differentiated by Sex and matched to the sex-specific plot.
The visualization can be exported to SVG, PNG or CSV formats, using the export buttons included at the bottom left of the configuration area.
Histo Barchart
The Histo Barchart visualization is an excellent tool for analyzing and visualizing incidence of histopathology findings in control animals. To do this, first filter for just control animals under Data Filters (DOSECAT = C). From there, select the tissue(s) and finding(s) of interest to display a bar chart by Species that shows the number of findings/total tissues example, colored by finding severity by default. The Y axis can be updated to # of Subjects with Findings or # Findings/Examined Subjects.
Additional filters can be added to further refine the results by sex, strain, test facility, vehicle, route, etc. The visualization can be exported to SVG, PNG or CSV formats, using the export buttons included at the bottom left of the configuration area.