MercatorR

MercatorR performs visualization and clustering of distance matrices, particularly from binary data, to enable unsupervised pattern discovery in large-scale biological datasets.

Key Features:

• Comprehensive visualization capabilities: Generates multiple visualizations using standard algorithms to compare structure and clustering results across analyses.
• Flexible distance metric selection: Supports a wide array of distance metrics for binary and other data types to highlight different aspects of similarity between observations.
• Facilitation of unsupervised analysis: Integrates distance-matrix computation with visualization and clustering to identify patterns and groupings without prior labels.

Scientific Applications:

• Genomics: Exploration of genomic binary data and detection of sample groupings or population structure.
• Proteomics: Identification of protein-level patterns and clustering based on similarity metrics derived from binary or presence/absence data.
• Ecology: Analysis of ecological presence/absence or binary trait datasets to reveal community structure and associations.
• Pattern recognition in large-scale datasets: Comparative assessment of distance metrics and visualizations to uncover structure in diverse biological datasets.

Methodology:

Calculates pairwise distances between observations using selected metrics and applies clustering algorithms to the resulting distance matrices to generate visualizations of patterns and structures.