Fig 1.
MicrobeTrace accepts input data in a variety of formats.
This figure displays the most common use cases and their required files.
Fig 2.
MicrobeTrace excels at rendering pathogen genetic distance networks and mapping visual characteristics to user-provided metadata.
(2A) The HIV-1 partial polymerase (pol) distance network, with a genetic distance threshold (d) of 1.5%. (2B) The same HIV-1 pol network shown in 3A with node positions held constant, but with a more stringent genetic distance threshold (d) of 0.5%. (2C) The same HIV-1 pol network shown in 3A with node positions held constant. Nearest connected neighbor links have been superimposed as dashed lines. The transparency of links that do not connect nearest neighbors has been increased. Gender and transmission risk factors have been mapped to node shape and color, respectively.
Fig 3.
MicrobeTrace allows the creation of informative dashboard visualizations by tiling of different views within the browser window.
(3A) Reports of high-risk contact between COVID-19 cases in clusters of size N ≥ 5, nodes are (i) colored by province, (ii) shaped by gender, and (iii) labeled with the total number of high-risk contacts. (3B) Geospatial map of clusters of size N ≥ 5 zoomed to show only Seoul, South Korea. (3C) Geospatial map of clusters of size N ≥ 2. Node positions have been randomly altered with the ‘jitter’ functionality to preserve patient privacy. (3D) In-application color and shape keys that offer interactive color-pickers and labeling. (3E) Incidence curve showing confirmed test date. Map tiles provided by http://stamen.com/ under CC BY 3.0; data by openstreetmaps.org.
Fig 4.
MicrobeTrace visualization does not require genomic or contact tracing data and can easily calculate aggregation and cross-tabulation tables in addition to visualizing histograms, alluvial/flow diagrams and Gantt charts.
Each diagram has an inset settings menu that describes the settings changes necessary to achieve them. (4A) City-level aggregation achieved via a single dropdown selection. (4B) Alluvial diagram of associations between the Type of Exposure to COVID-19, Province, and Symptom Onset Date. (4C) Gantt charts to describe the span of time between Symptom Onset, Positive Test Confirmation, and Hospital Release Date. (4D) Age histogram, binned by decade and colored by gender. This histogram illustrates a trend identified early in the Korean COVID-19 outbreak, wherein a disproportionate number of middle-age female cases were diagnosed (4E) Cross-tabulation table of cases by City and Age categories.
Fig 5.
MicrobeTrace’s primary users are public health officials with the bulk of usage during the work week, as opposed to during the weekend.
In orange, are the number of monthly weekday users. In blue, are the number of monthly weekend users. Each month’s mean daily user count is mapped and colored by day of the week. A lineal regression for each day type is shown to smooth the month-to-month effects and highlight the increasing usage trend. (5A) Monthly user traffic, stratified by weekday and weekend use. (5B) Monthly user traffic categorized by CDC and non-CDC traffic and stratified by weekday and weekend use.