Multi-omic and quantum machine learning integration for lung subtypes classification

doi:10.1016/j.future.2025.107905

Multi-omic and quantum machine learning integration for lung subtypes classification

2025-06-05

0

PRE

AI

M

Mandeep Kaur Saggi

A

Amandeep Singh Bhatia

DOI：10.1016/j.future.2025.107905

Original

Original request for help

Save

Abstract

En 中文

• This study introduces MQML-LungSC, a novel Quantum Machine Learning framework that integrates DNA-me, RNA-seq, and miRNA-seq multi-omics modalities for accurate lung cancer subtype classification. • A statistical test is applied to multi-omics data for first-level feature extraction to find the significant/insignificant features of subsets. • Feature selection is performed on single-omics modalities, with comparisons to classical ML models for evaluating importance and interpretability. • Hybrid Quantum-Classical Neural Network models variants (256, 64, 32) reveal key subtype-specific diagnostic biomarkers across LUAD and LUSC cohorts. • Advanced QNNs tuning and multi-omics analysis enable next-gen QML workflows for precise biomarker discovery, precision diagnosis, prognosis, and personalized cancer care.

Keywords:

Quantum computing

Quantum machine learning

Multi-omic

Biomarkers

LUAD

LUSC

Multi-omic and quantum machine learning integration for lung subtypes classification

Abstract

Journal

Organization