Related Experiment Video
Updated: May 25, 2026
07:47
Multiomics Analysis of TMEM200A as a Pan-Cancer Biomarker
Published on: September 15, 2023
Interactome-transcriptome integration for predicting distant metastasis in breast cancer.
Maxime Garcia1, Raphaelle Millat-Carus, François Bertucci
1Aix-Marseille Univ, F-13284 Marseille, France. maxime.garcia@inserm.fr
Bioinformatics (Oxford, England)
|January 13, 2012
Summary
A new algorithm, interactome-transcriptome integration (ITI), improves genomic signature stability and generalization for predicting breast cancer metastasis. This method enhances accuracy compared to existing approaches.
Area of Science:
- Bioinformatics
- Genomics
- Cancer Research
Background:
- High-throughput gene expression profiling generates genomic signatures for predicting patient outcomes.
- Current genomic signatures face limitations like training set dependency and poor generalization.
Purpose of the Study:
- To develop a novel algorithm for more robust and generalizable genomic signatures.
- To improve the prediction of distant metastasis in breast cancer.
Main Methods:
- Introduced the interactome-transcriptome integration (ITI) algorithm.
- Integrated large-scale protein-protein interaction data with gene expression datasets.
- Applied ITI to estrogen receptor-specific breast cancer data.
Main Results:
- ITI-derived signatures demonstrated improved stability (11-35%) and generalization on independent data.
- Achieved higher prediction accuracy (53-74%) compared to previous methods.
- Successfully extracted estrogen receptor-specific genomic signatures.
Conclusions:
- The ITI algorithm offers a more stable and generalizable approach for genomic signature extraction.
- This method enhances the prediction of distant metastasis in breast cancer.
- ITI represents a significant advancement over existing signature prediction techniques.
