An open-source R/Bioconductor package for tabular omics data analysis using a supra-hexagonal map

News

• The artwork called supraHex has won The Best Artwork Award in ISMB 2014. This artwork is automatically done and is reproducible here.
• Demonstrated in a variety of genome-wide datasets such as:

  leukemia patient transcriptome, time-course transcriptome, RNA-seq data, DNA replication timing, cell type evolution, and clonal population structure.

Introduction

The supra-hexagonal map is a giant hexagon on a 2-dimensional map grid seamlessly consisting of smaller hexagons.

supraHex intends to meet the need for quickly understanding genome-wide biological data, which usually involve a large number of genomic coordinates (e.g. genes) but a much smaller number of samples.

supraHex first uses a supra-hexagonal map to self-organise the input omics data, and then post-analyses the trained map for integrated tasks: simultaneous analysis of genes and samples, and multilayer omics data comparisons.

supraHex aims to deliver an eye-intuitive tool and a dedicated website with extensive online documentation and easy-to-follow demos.

For more, see slides and poster in ISMB2014.

Features

• The supra-hexagonal map trained via a self-organising learning algorithm;
• Visualisations at and across nodes of the map;
• Partitioning of the map into gene meta-clusters;
• Sample correlation on 2D sample landscape;
• Overlaying additional data onto the trained map for exploring relationships between input and additional data;
• Support for heatmap and tree building and visualisations;
• Used by the package dnet for network-based sample classifications;
• This package can run on Windows, Mac and Linux.

Workflow