CMcrpred
CMcrpred predicts risk in cutaneous melanoma patients by integrating clinico-pathological features and gene-expression profiles to improve prognostication.
Key Features:
- Integration of Clinico-Pathological and Gene-Expression Data: Combines clinico-pathological features such as Breslow thickness and AJCC tumor staging with gene-expression profiles associated with cancer-related pathways.
- Biomarker Development: Utilizes biomarkers derived from clinical data and gene-expression profiles, with emphasis on pathways including apoptosis and Notch signaling, for prognostic assessment.
- Gene-Expression Based Models: Gene-expression models show hazard ratios (HR) up to 2.52 for the apoptotic pathway and up to 2.57 when apoptotic genes are combined with Notch pathway genes.
- Clinico-Pathological Feature Models: Individual clinical features, for example Breslow thickness, produced a maximum HR of 2.45.
- Combined Models: Models integrating clinical variables and gene-expression data achieved a maximum HR of 3.19.
- Ensemble Method for Clinical Variables: An ensemble method using only clinico-pathological features produced a maximum HR of 6.40, exceeding performance of gene-expression models and AJCC staging.
Scientific Applications:
- Melanoma prognostication and risk stratification: Provides quantitative risk estimates to support prognostic assessment in cutaneous melanoma patients.
- Biomarker evaluation and pathway analysis: Supports identification and evaluation of prognostic biomarkers and the involvement of pathways such as apoptosis and Notch signaling in melanoma progression.
Methodology:
Analyzing gene-expression profiles related to key cancer pathways, evaluating clinico-pathological features for prognostic value, creating models that integrate both data types, and implementing an ensemble method focused solely on clinical variables.
Topics
Details
- Added:
- 1/18/2021
- Last Updated:
- 2/12/2021
Operations
Publications
Arora C, Kaur D, Lathwal A, Raghava GP. Risk prediction in cutaneous melanoma patients from their clinico-pathological features: superiority of clinical data over gene expression data. Heliyon. 2020;6(8):e04811. doi:10.1016/j.heliyon.2020.e04811. PMID:32913910. PMCID:PMC7472860.