Find in Library

Summary: A regularly shaped grid is useful for analyzing data particularly at multilayer levels, where patterns can be visually represented and analytically compared—conceptually similar to Picasso's cubism. Here we introduce ATLAS, featuring a suite of spatially ordered maps designed for representation and comparison of patterns seen in regulatory genomic data. It produces a landscape learned from input data and enables landscape-guided correlation with additional data. We illustrate its use for multilayer data comparison on the same cell type, and for comparisons involving different cell types, revealing information in a scientifically insightful and also visually intuitive way. The data-driven and visual-aided ability of ATLAS presents a general strategy for regulatory genomic data analysis. : Cubism; Regulatory Genomics; Artificial Intelligence Subject Areas: Cubism, Regulatory Genomics, Artificial Intelligence