MulMarker: a comprehensive framework for identifying multi-gene prognostic signatures

TL;DR

MulMarker is a comprehensive framework that efficiently identifies multi-gene prognostic signatures, aiding in patient risk stratification and survival prediction across various diseases.

Contribution

We introduce MulMarker, a novel framework with modules for identifying multi-gene signatures, including a chatbot, that improves prognostic analysis and patient stratification.

Findings

Identified a cell cycle-related prognostic signature with multiple CCN genes.

Successfully stratified patients into high- and low-risk groups with significant survival differences.

Released the MulMarker code for public use.

Abstract

Prognostic signatures play an important role in clinical research, offering insights into the potential health outcomes of patients and guiding therapeutic decisions. Although single-gene prognostic biomarkers are valuable, multi-gene prognostic signatures offer deeper insights into disease progression. In this paper, we propose MulMarker, a comprehensive framework for identifying multi-gene prognostic signatures across various diseases. MulMarker comprises three core modules: a chatbot for addressing user queries, a module for identifying multi-gene prognostic signatures, and a module for generating tailored reports. Employing MulMarker, we identified a cell cycle-related prognostic signature that consists of CCNA1/2, CCNB1/2/3, CCNC, CCND1/2/3, CCNE1/2, CCNF, CCNG1/2, and CCNH. Based on the prognostic signature, we successfully stratified patients into high-risk and low-risk groups. Notably, our analysis revealed that patients in the low-risk group demonstrated a significantly higher survival rate compared to those in the high-risk group. Overall, MulMarker represents an efficient approach for the identification of multi-gene prognostic signatures. We release the code of MulMarker at https://github.com/Tina9/MulMarker.