Normalization and Selecting Non-Differentially Expressed Genes Improve Machine Learning Modelling of Cross-Platform...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Quantitative analysis of biological processes requires robust normalization.
• Cross-platform integration of RNA microarray and RNA-sequencing (RNA-seq) data for machine learning (ML) is challenging due to lack of independent validation.
• Improving ML model performance on independent datasets across platforms is crucial.

Purpose of the Study:

• To test the hypothesis that non-differentially expressed genes (NDEG) can improve transcriptomic data normalization and cross-platform ML modeling.
• To evaluate NDEG-based normalization for classifying breast cancer molecular subtypes using independent microarray and RNA-seq datasets.

Main Methods:

• Used TCGA breast cancer microarray and RNA-seq datasets as independent training and test sets.
• Selected NDEG (p > 0.85) and differentially expressed genes (DEG) (p < 0.05) based on ANOVA p-values.
• Applied NDEG for normalization and DEG for classification in ML models, testing cross-platform performance.

Main Results:

• NDEG and DEG selection significantly improved ML model classification performance.
• Nonparametric statistical normalization methods outperformed parametric methods.
• LOG_QN and LOG_QNZ normalization with neural networks showed superior performance.

Conclusions:

• NDEG-based normalization is effective for cross-platform ML model testing on independent datasets.
• This approach shows promise for improving transcriptomic data analysis and classification.
• Further research is needed to validate NDEG normalization across diverse datasets and omics types.