Welcome to SeptiSearch! Here you can browse, explore, and download curated molecular results derived from transcriptomic sepsis studies. The database and app currently catalogs over 20,000 unique molecules from more than 70 publications.

To get started, select one of the tabs at the top of the page, described below:

  • Explore the Database makes it easy to search all curated gene sets by keyword or molecule, filter with select criteria, and view all the molecules in a gene set
  • Visualize the Database plots the most common molecules, and provides filters to see which occur most frequently based on a number of attributes
  • Perform Pathway Enrichment allows users to upload their own list of genes, or use one of the curated sepsis gene sets, to test for enriched pathways/terms using ReactomePA and enrichR
  • Test for Enriched Sepsis Gene Sets can be used to test your own expression data (e.g. counts from RNA-Seq) for dysregulation of the curated sepsis gene sets

SeptiSearch was created by Travis Blimkie, Jasmine Tam & Arjun Baghela from the Hancock Lab at the University of British Columbia. If you'd like to learn more, or to report bugs or issues, click the button below to visit our About page.



SeptiSearch is a Shiny app in which you can browse, explore, and download curated molecular gene sets derived from transcriptomic sepsis studies. The app currently allows access to over 20,000 unique molecules from over 70 publications. It was created by Travis Blimkie, Jasmine Tam & Arjun Baghela from the Hancock Lab at the University of British Columbia. The last update to the data was performed in November 2022. Travis is the main developer for the Shiny app and handles maintenance & updates. Jasmine performed all the signature curation from datasets in peer-reviewed research articles and publicly available pre-prints. Arjun served as the supervisor for the project.

Gene Sets (i.e. the Gene Set Name column) are defined based on a number of columns/fields from each study, such that one study may have multiple gene sets. For example, if one study compares two groups of sick patients (e.g. severe and mild sepsis) to the same group of healthy controls, that study would have two gene sets. The fields used to determine the Gene Sets are: Timepoint, Case and Control Condition, Tissue, and Gene Set Type.


A tutorial is available which provides detailed insturctions for using SeptiSearch and its different functions, hosted on the GitHub repository: https://hancockinformatics.github.io/SeptiSearchTutorial/

Reporting problems

If you encounter a problem or bug with the app, please submit an issue at the Github page. Include with your issue details on the problem so we can reproduce it, and any inputs if relevant (e.g. your list of genes submitted to the Perform Pathway Enrichment tab).


We would like to acknowledge and thank the Canadian Institutes of Health Research (CIHR) for providing the funding for this project.

Database details & app methods

'Tissue Class' definitions

The Tissue Class column contains a controlled vocabulary to describe the type of tissue in which a study was performed: 'Blood' denotes sampling was done on whole blood; 'Blood Cells' indicates a specific cell type was isolated from blood; 'Lung Cells' means particular cells or tissue samples were extracted from the lung; 'Other Cells' is used to describe any remaining entries not fitting those mentioned previously. We recommend checking the original source if you require more detailed information.

Perform Pathway Enrichment

Input gene mapping between ID types is performed using data obtained via the biomaRt package. Biological pathway/term enrichment is performed using ReactomePA and enrichR. The following resources are searched using enrichR: MSigDB's Hallmark collection, and the three main GO databases (Biological Process, Cellular Component & Molecular Function). For both methods, the results are filtered using an adjusted p-value threshold of 0.05.

Test for Enriched Sepsis Gene Sets

Gene Set Variation Analysis is performed using the GSVA package, and the heatmap visualization is created with pheatmap. Specified parameters include the gsva method and a Gaussian kernel. Genes with zero variance across all samples are removed prior to the analysis. Example data for GSVA represents a subset of the GEO record GSE65682.


SeptiSearch is written in R, and uses the following packages & resources:

Create beautiful web apps with R.
Extend Shiny functionality using JavaScript.
A suite of packages for data manipulation.
An R interface to the DataTables JavaScript library.
Interactive plots and visualizations.
An R package to access BioMart databases.
Perform pathway enrichment analysis using Reactome data.
Access the Ma'ayan Lab's gene set enrichment services from R.
Gene Set Variation Analysis for microarray and RNA-Seq data.
Easy and robust heatmap visualizations.