EnMCB

EnMCB builds signatures based on DNA methylation correlated blocks for survival prediction by combining multiple machine learning models in a stacked ensemble framework.

Key Features:

• Methylation Correlated Block Identification: The package automatically partitions the genome into blocks of tightly co-methylated CpG sites, termed methylation correlated blocks.
• Ensemble Model Construction: Cox regression, support vector regression, mboost, and elastic-net models are combined in the ensemble model.
• Survival Prediction: Diagnostic capacities for predicting patients' survivals are generated after partitioning and modeling.
• Differentially Methylated Block Prediction: Machine learning models can be constructed to predict differentially methylated blocks.

Scientific Applications:

• Cancer Diagnosis: The combination of multiple CpGs into a single block may promote cancer diagnosis based on contiguous cytosine-phosphorothioate-guanine (CpG) sites in the same DNA strand that may be modified by a methyltransferase or demethylase together.
• Disease Progression Prediction: Models predict disease progression based on methylation correlated blocks using ensemble models.
• Survival Analysis: The tool provides diagnostic capacities for predicting patients' survivals using methylation profiles from The Cancer Genome Atlas.

Methodology:

The package automatically partitions the genome into methylation correlated blocks, then combines Cox regression, support vector regression, mboost, and elastic-net models in a stacked ensemble for survival prediction.