Predicting Cancer Treatment Response with Metagenes
Author Information
Author(s): Li Qiyuan, Eklund Aron C, Birkbak Nicolai J, Desmedt Christine, Haibe-Kains Benjamin, Sotiriou Christos, Symmans W Fraser, Pusztai Lajos, Brunak Søren, Richardson Andrea L, Szallasi Zoltan
Primary Institution: Center for Biological Sequence Analysis, Technical University of Denmark
Hypothesis
Can consistent metagenes derived from cancer expression profiles predict chemotherapy response?
Conclusion
The study found that metagenes derived from multiple data sets can effectively predict chemotherapy response in breast cancer and other tumor types.
Supporting Evidence
- The study derived four metagenes from five cohorts of double-negative breast cancer.
- Three of the metagenes showed strong associations with chemotherapy response.
- The method was also applied to ovarian and lung cancer, yielding predictors of survival.
Takeaway
Scientists found a way to use patterns in cancer genes to help doctors know which treatments will work best for patients.
Methodology
The study used unsupervised methods to derive metagenes from expression profiles of various cancer types and validated their predictive power in independent cohorts.
Potential Biases
Potential overfitting due to the use of multiple data sets without independent validation for all findings.
Limitations
The study's findings may not be applicable to all cancer types due to the specific focus on certain cohorts.
Participant Demographics
The study included various cohorts of breast cancer patients, specifically focusing on double-negative breast cancer.
Statistical Information
P-Value
0.005
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website