Expression

Overview: The expression landscape of 3432 RNA-Seq fresh frozen tumor samples (1389 blood tumors, 888 solid tumors, and 1155 brain tumors) in St. Jude Cloud is displayed via a t-SNE plot (Figure 1) generated using the St. Jude Cloud RNA-Seq Expression Analysis workflow.

Figure 1: tSNE for Blood, Brain, and Solid Samples. Metadata details for each sample can be accessed by mousing over the data points. This visualization is supported by D3.

Features: A user can explore across the 3 tSNE plots: Blood, Solid, or Brain tumor tabs and employ the features listed below:

Subtype categorization- Subtypes are denoted by a specific color and a subset have been labeled on the plot.

Sample Summary- A user can select a data point on the plot that opens a sample summary drawer annotating relavent metadata and information.

Filters - Filters are organized by Tumor Sample, Patient Phenotype, and Sample Preparation. Once a filter is selected, the subtype labels will be disabled. Functionality of each is further described below.

Sample Search - A user can search individual sampleIDs or bulk IDs that are comma separated. The sampleIDs must be exact and cannot be fuzzy searched.

Lasso - Allows a user to select a specific region on the plot to retrieve a list of samples to enable further investigation. To view the sample summary of the lassoed samples, click the “Data” icon in the top right of the subnavbar. See GIF below.

Pan/Zoom - Allows a user to examine regions of the plot in more detail, this disables any labels.

Filters Explained

Tumor Sample

1. Sample ID - A user can search individual St. Jude CompBio IDs or bulk search IDs that are comma separated. This field allows for a multi-select.
2. Subtype - This is a modal whereby a user can custom select which subtypes to view in the plot. Child nodes will automatically become enabled or disabled if a parent node is (de)selected. The number of samples and the subtype color is desginated in the modal for reference.
3. Subtype Biomarker - This field allows a multi-select of subtype biomarkers to be applied to the plot. Note: the user cannot apply a general gene like “CTNNB1” to be applied across the plot. The user must select all biomarkers they are interested in seeing from the dropdown
4. Sample Type - This field is a multi-select dropdown.

Patient Phenotype

1. Sex - A multi-select dropdown.
2. Age at Diagnosis - A scale whereby a user can manually type in the age parameters or use the scale (in years). A user can type in any age, even passed our “35+” parameter.
3. Race - This is a multi-select dropdown.
4. Ethnicity - This is multi-select dropdown.

Sample Preparation

1. Library Selection Protocol - This is a multi-select dropdown.
2. Preservative - This is a mutli-select dropwdown