HDMAC: A Web-Based Interactive Program for High-Dimensional Analysis of Molecular Alterations in Cancer

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• High-throughput genomic technologies generate vast datasets requiring advanced statistical tools.
• Identifying molecular alterations is crucial for personalized medicine, prognostic biomarkers, and drug targets.

Purpose of the Study:

• To develop a user-friendly, web-based platform (HDMAC) for high-dimensional analysis of molecular alterations in cancer.
• To provide statistical tools for identifying potential prognostic biomarkers and drug targets.

Main Methods:

• Developed HDMAC, a web platform offering penalized regression models (Ridge, Lasso, adaptive Lasso), Cox regression, and logistic regression.
• Incorporated a first-step screening to address multiple-comparison issues in large genomic datasets.
• Utilized cross-validation for prediction power estimation and provided R codes for TCGA data download.

Main Results:

• HDMAC was illustrated using gene mutation and mRNA expression data from ovarian and bladder cancer patients.
• Analysis identified candidate genes associated with mutations or abnormal gene expression in these cancers.
• The platform facilitates rigorous analysis and validation of high-dimensional genomic data.

Conclusions:

• HDMAC provides a valuable solution for analyzing complex cancer genomic data.
• The platform supports the identification of candidate genes for further investigation in cancer research.
• It aids in advancing personalized precision medicine through robust data analysis.