Predictive Gene Lists for Breast Cancer Prognosis
Author Information
Author(s): Sivaraksa Mingmanas, David Lowe
Primary Institution: Aston University
Hypothesis
Can patient-specific gene expression profiles be used to distinguish between good and poor prognosis in breast cancer patients?
Conclusion
The study concludes that many patients are unclassifiable based on predictive gene lists due to uncertainty in gene expression profiles.
Supporting Evidence
- Small subsets of patient-specific predictive gene lists have insufficient prognostic dissimilarity.
- Uncertainty across multiple gene expressions prevents confident patient grouping.
- Comparative projections across different gene lists yield similar results.
Takeaway
This study shows that using a small number of genes to predict breast cancer outcomes can be confusing, and some patients can't be easily classified into good or bad groups.
Methodology
The study used nonlinear topographic projection maps based on inter-patient gene-list dissimilarities to visualize and classify breast cancer prognosis.
Potential Biases
The study acknowledges potential biases due to the small number of patients and the high dimensionality of gene expression data.
Limitations
The study highlights the limitations of small sample sizes and the random correlation of gene lists with outcomes.
Participant Demographics
The study involved 78 sporadic lymph-node negative breast cancer patients, with 34 developing metastases and 44 remaining cancer-free within 5 years.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website