ACES

ACES performs clustering and visualization of large-scale genomic and epigenomic datasets to stratify samples and link genetic or epigenetic variants, including DNA methylation and RNA-Sequencing data, to phenotypes.

Key Features:

• Integrative Clustering and Visualization: Integrates clustering algorithms with dimensionality reduction and visualization techniques, including Principal Component Analysis (PCA) and heat maps, to provide 2D and 3D views of samples.
• Phenotype Mining and Cluster Enrichment: Automatically mines phenotype lists to identify cluster enrichment and relationships between molecular markers and complex phenotypes.
• Cluster Number and Boundary Estimation: Estimates the number of clusters and their boundaries using an internal method to improve stratification accuracy.
• Application to Genomic Data: Analyzes pre-filtered DNA methylation and RNA-Sequencing datasets to link molecular markers to phenotypic outcomes.
• Individual-level Stratification: Stratifies individuals or samples to support individual-level predictions based on molecular data.

Scientific Applications:

• Variant-Phenotype Association: Identification of genetic or epigenetic variants associated with specific traits or diseases through cluster-based enrichment analysis.
• Outlier Detection: Detection and characterization of outlier samples within large genomic and epigenomic cohorts.
• Personalized Medicine and Genomic Epidemiology: Support for studies linking molecular marker patterns to individual phenotypes for personalized medicine and population-level genomic epidemiology.
• Systems Biology: Exploration of molecular marker relationships and phenotype stratification to inform systems-level biological hypotheses.

Methodology:

Integrates clustering algorithms with visualization techniques (PCA and heat maps), performs phenotype mining for cluster enrichment, and estimates cluster numbers and boundaries to enable sample stratification.