🗒️Generate Reports

How to leverage you graph models to generate reports for the purposes of target prioritization, indication selection and more.

Use the top application toolbar and select “Graph Models". Use the left-hand panel to select a specific graph to generate a report. Once all of your lines of evidence have been created and the graph has been updated, you can create your first report.

Graph reports can be used to prioritize and rank any set of objects within your Knowledge Graph. In this example, we will be creating a report that will prioritize genes that would be suitable targets for renal cell carcinoma.

To create a report, you will need to specify the following

• Name

  • Short description of the report

• Description

  • Longer, detailed description of the report

• Concept A Value

  • This refers to a specific object associated with the concept

  • Example: The graph model is Disease to Gene. Concept A is a Disease, so we will select the disease of interest, "renal cell carcinoma".

• Concept B Value

  • This refers to a specific object associated with the concept

  • Example: If your graph model is Disease to Gene and Concept B is Gene, we will leave this input blank so that it is unbiased.

• Only 1 concept can be unbiased. The objects associated with the concept that is unbiased will be used for ranking and prioritization within the report. Please note that the inputs are case sensitive. If no results are returned in the report there may be no observations for that object, e.g. a specific disease.

Report Results

At the top of the report, you will find all of the report metadata and a summary. Objects are listed in ranked order based on their overall score. Scores are computed based on graph model evidence, available data, and scoring formulas.

On the left of the report, you will find the lines of evidence (scientific criteria you have defined in your model).

Exploring the Report

Use the right and left arrows to quickly browse across all of the objects returned in the report.

Selecting a data point will enable you to explore the associated evidence identified across the public and private data within your graph.

Scoring Data Vs Graph Paths

When a data point is selected, the data from your graph will be displayed. The first tab displays the specific relationship or concept used to compute the overall score for that line of evidence.

The second tab will display all of the data associated with the graph path. The specific data properties and the order in which they are displayed can be customized in the graph path table.

AI Summaries & Dashboards

Hover over each object at the top of the report, and you will see a tool tip that says "Summary". When selected, it will generate an AI summary that takes into taking into consideration the lines of evidence in your graph and the available data. Below the AI summary, you will find a pre-populated dashboard that displays the data points associated with each line of evidence.

Evidence Requirements

The check box next to each line of evidence can be used to filter and further prioritize data points. The check boxes can operate under "AND" or "OR" inclusion criteria. When using "AND" inclusion criteria, this will filter for objects that have all of the selected lines of evidence. When using "OR" inclusion criteria, this will filter for objects that have at least 1 of the selected lines of evidence. By default, the settings operate under "AND" inclusion criteria; this can be updated by selecting the "Inclusion Criteria" button located in the top left hand of the report.

Export Report Data

To export a list of objects returned within your report, select "Export Objects" from the top right hand corner of the report. Specify the start and end limit for the number of objects to be returned and a CSV will be exported to your local machine.