Data

System Pro uses AI to extract findings from peer-reviewed studies and other databases (the performance of our AI models can be found here) and structures them for further analysis and synthesis. With this view, you now have access to all of this structured data to supercharge your literature reviews and meta-analyses.

Types of findings

Using the View function nested under the tabs, you can toggle between two types of findings: Statistical and Mechanistic.

Statistical findings are any relationship reported in the literature with two variables, a statistic type, a measured statistic value, and a confidence level (either p-value and/or confidence interval). These findings are your building blocks for meta-analysis.

Mechanistic findings do not have any associated statistics, and instead are validated interactions. They contain two agents (e.g., a gene and a protein) as well as a mechanism type. These findings are your building blocks for understanding pathways.

Filtering and exploring the data

Data can be filtered by variable names on the "statistical findings" view and agent names on the "mechanistic findings" view. You can use these filters to identify all variables or agents that contain the same string of text.

For mechanistic findings, you can also hover over the agent names to view and access each agent's external identifiers.

Selecting and exporting data

When you want to extend your analysis, the data is available to export in its structured form using the "Export" button in the upper right of the table. As on the Studies tab, you can export data as either a csv file or an RIS file.

Suggesting revisions

System places a high value on the trustworthiness of information and employs many measures to ensure the accuracy of our data. Yet, error is still possible and we invite our community of users to help us identify these errors by suggesting a revision.

This can be done by scrolling to the far right of the table to select “Suggest revision”, then providing details on the issue you encounter. Your suggestions will be sent to our internal team for triaging. While under review, the finding will be marked with “Pending Review”, visible to all users, and it will be excluded from any synthesis. Once the review is complete, you will receive an email with the results of the review and the data fixed as needed.