Enhanced prognostic signature for lung adenocarcinoma through integration of adjacent normal and tumor gene expressions
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.Combining tumor and normal tissue gene expression data improves lung adenocarcinoma prognostic signatures. A combined model demonstrated superior generalizability and clinical relevance compared to models using only tumor or normal tissues.
Area Of Science
- Oncology
- Bioinformatics
- Genomics
Background
- Cancer prognosis signatures traditionally use tumor or normal tissue gene expression.
- The benefit of combining both tissue types for prognostic signatures is unexplored.
Purpose Of The Study
- To develop and evaluate prognostic models for lung adenocarcinoma (LUAD) using tumor, normal, and combined gene expression profiles.
- To assess the generalizability and clinical significance of these models.
Main Methods
- Developed three LUAD prognostic models using The Cancer Genome Atlas (TCGA) data: tumor-derived, normal-derived, and combined (COM).
- Validated models on TCGA and three independent datasets, analyzing overall survival, discrimination, and calibration.
- Assessed clinical significance using clinical features and GSE229705 validation.
Main Results
- The combined (COM) model showed the best performance on independent datasets and superior overall survival prediction.
- COM-derived risk scores demonstrated the strongest clinical significance and generalizability.
- Tumor-derived models showed potential overfitting on the primary dataset; normal-derived models performed poorly.
Conclusions
- Combining tumor and normal gene expression data enhances prognostic signature performance and generalizability for LUAD.
- The COM model offers a more clinically relevant and robust approach to cancer prognosis.
These videos have been matched automatically. Contact us if they are not relevant.
These videos have been matched automatically. Contact us if they are not relevant.
