A Heuristic Algorithm for Identifying Molecular Signatures in Cancer.

Area of Science:

• Bioinformatics
• Computational Biology
• Cancer Research

Background:

• Identifying cancer molecular signatures (genes, microRNAs) is crucial for early detection.
• Challenges exist in analyzing high-dimensional omics data (gene/miRNA expression) with limited samples.

Purpose of the Study:

• To develop a heuristic algorithm, HAMS, for identifying robust molecular signatures for cancer diagnosis.
• To address the challenge of high dimensionality in gene and microRNA expression datasets.

Main Methods:

• Modeled cancer molecular signature identification as a multi-objective optimization problem.
• Proposed HAMS, a heuristic algorithm with an elitist-guided individual update strategy.
• Utilized gene-expression and miRNA-expression datasets for evaluation.

Main Results:

• HAMS identified a small set of molecular signatures highly relevant to cancer.
• The algorithm demonstrated superior performance compared to seven state-of-the-art methods.
• Obtained molecular signatures were validated for biological significance.

Conclusions:

• HAMS effectively identifies biologically significant cancer molecular signatures from complex omics data.
• The algorithm offers an improved approach for cancer diagnosis by reducing signature redundancy.
• HAMS shows promise for advancing precision oncology and biomarker discovery.