About IOSig


Immuno-Oncology Signatures Explorer (IOSig) is an interactive Shiny application meant to facilitate the investigation of immune checkpoint inhibitor (ICI) treated datasets with gene expression biomarker signatures for prediction of responder / non-responder. Previously published gene expression signatures have been collected in addition to publicly available ICI treated RNA-seq cohorts to allow for query their own datasets or signatures against. In addition, it also possible to explore the data within the application without providing your own.

IOSig Methods

User Guide
Dataset Analysis
The Dataset Analysis tab is meant for users to upload their own RNA-seq expression datasets coupled with the necessary clinical data. The RNA-seq data is expected to be in a comma or tab separated format with the columns containing the sample ids and rows containing genes. The genes should be in HUGO format. Additionally the samples cannot start with a number. If your samples start with a numerical value, add a character, like 'X' at the start.
It is also necessary for the uploaded clinical data to have specific column names. These column names are: 'Sample_ID' (sample name), 'response' (ICI response), 'os' (overall survival time), and 'os_event' (overall survival event). 'reponse' and 'os_event' should be binary columns where 1 represents responder to ICI treatment and censoring in os_event.
Example RNA-seq and Clinical files can be downloaded from the left side panel on the Dataset Analysis page. Additionally, the datasets used for comparison in the AUROC portion can be filtered and selected in the left side panel.
User Signature Analysis
The User Signature Analysis tab allows a user to upload a list of genes in HUGO format to query against the datasets within IOSig. The genes should be separated by a new line and pasted into the left side panel. A unique name for the uploaded gene signature is also required. An example gene signature file can be downloaded from the left side panel as well.
Published Signature Analysis
The 'Published Signature Analysis' tab is similar to the 'User Signature Analysis' tab except that you can select from a variety of previously published gene signatures to query the IOSig datasets.

AUROC Analysis

Area under the receiver operating characteristic curve (AUROC) is a performance measurement metric for a binary classifier. In IOSig, AUROC is utilized to assess the ability of gene signature to distinguish responder and non-responder based on the gene expression values for a sample. The curve is plotted with the true positive rate on the y-axis and false positive rate on the x-axis. A higher area under the curve (AUC) value the better the gene signature is at predicting responder / non-responder. An AUC value of 0.5 signifies the gene signature has no ability to separate responder / non-responder while 1 is a perfect classifier.
Three AUROC analyses are calculated:
  • AUROC Overview
    • The individual AUROC plots for each dataset / signature (depending on analysis method).
  • AUROC Comparison
    • A graphical representation of the AUROC values of the selected datasets over all signatures.
  • AUROC Correlation
    • A correlation heatmap created from the AUROC values of all gene signatures.

Mann Whitney

The Mann Whitney U test provides a way to test for differences between two groups (responder, non-responder). In this case, the metric used to compare responders and non-responders is the score calculated for each sample from each gene signature. For this analysis, a box plot to demonstrate the distribution of responder and non-responder samples is generated in addition to a table with basic summary information. For uploaded datasets, response information should have a column title of 'response' with 1 signifying responder and 0 signifying non-responder.

Survival

A Survival analysis is available for datasets that have the necessary included information. Using the overall survival information, a Kaplan-Meier graph is generated. When uploading new clinical data, the column representing time should be 'os' with the column signifying an event labeled 'os_event'.

Gene Set Enrichment Analysis

Gene Set Enrichment Analysis (GSEA) is a method which analyzes gene expression data and focuses on finding changes at the level of gene sets as opposed to individual genes. This allows for the detection of changes which may be insignificant at the gene level, but significant across a summation of genes that encompass a biological process or network. Normalized enrichment score (NES) is the primary result of a GSEA analysis. This statistic reflects the level of which a gene set is overrepresented at the top or bottom of the ranked list of genes. In IOSig, the GSEA analysis and plots are generated with responders or high category samples at the top (dependent on dataset or signature analysis).


We would be grateful if you cite IOSig in your work:
Samuel Coleman, Caroline Wheeler, Rebecca Hoyd, Louis Denko, Ching-Nung Lin, Muhammad Zaki Hidayatullah Fadlullah, Siwen Hu-Lieskovan, Christine H. Chung, Ahmad A. Tarhini, Daniel Spakowicz and Aik Choon Tan, 2024
Version 1.62

Expression data included in this application:

Signature data included in this application:

Clinical Filter Table

Uploaded Expression Data

Uploaded Clinical Data

AUROC Overview

AUROC Comparison

AUROC Correlation


Survival



Mann Whitney U Test

GSEA

Enrichment Plots

User Data

Comparison Data

Clinical Filter Table

Signature Summary

AUROC Overview


AUROC Comparison


AUROC Correlation


Survival



Mann Whitney U Test


Cohort Specific Table


Signature Specific Table

GSEA

Enrichment Plots

User Data

Comparison Data

Clinical Filter Table

Signature Overview


Gene Overview


AUROC Overview

Download

AUROC Comparison

Download

AUROC Correlation


Survival

Download


Mann Whitney U Test

Download

Cohort Specific Table


GSEA

Download