Case Study Header
CASE STUDY

Survival (Time to Event)

Case Study part1
Applying just the basic Preliminary analysis, JADBIO analyzed the expression data from 621 miRNAs and 501 tumor profiles. The analysis involved the testing and optimizing of hyperparameters in 30 models in just 14 seconds. The resulting Best Performing model has a concordance Index of 0.768 with confidence intervals between [ 0.690, 0.842 ]. Kaplan-Meir visualizations split the patients into two, three, or more risk groups based on the 14 selected most predictive features. This automated analysis corroborates the findings of Forst’s publication, cited above.
Case Study Jadbio-icon-rgb

Reading the molecular labels in cancer

Predicting survival time in Low-grade glioma (LGG) patients is difficult. Current prognostic indicators, such as patient age, tumor size and location, do not explain the survival differences seen in LGG patients.

One of the unique capabilities of JADBIO is to predict Survival, even when samples are few, features are many, and the data is highly censored. All of these challenges are present in this study, which identifies miRNA expression levels that predict survival in LGG patients.
Low-Grade Gliomas
Case Study part2
DOWNLOAD CASE STUDY pdf_img DOWNLOAD TUTORIAL pdf_img SEE PROJECT

USE CASES