Scope and opportunity

Case study 1: TCGA Copy Number / Expression

Accessing 'Consortium' data

Representing integrated data sets

Key components

Samples

• Including 'phenotypic' attributes measured on each

Assays

• Often, rectangular features x samples matrix() of values
• A different representation: data.frame() of feature, sample, value tuples.

Experiments

• Sample inputs to each experiment
• Assay outputs from each experiment
• Additional experiment-specific information, each date on which samples were processed

Alternative Representations

R-based representations

Other approaches

• Data base
• 'Scientific' representations, e.g., HDF5
• 'Computational' representations, e.g., HDFS

A preliminary design

Statistical and other challenges

Practical