Figure 1.

Integrating biological data from multiple sources to construct regulatory network models. State-of-the-art computational techniques combine information from large-scale public data sets with omics measurements collected from the samples under study to generate multiscale causal models. Whenever possible, samples should be collected from multiple tissues or states in an individual (for example, diseased and healthy) at several time points. Measurements from single tissues miss important context-specific regulatory interactions that are responsible for disease. Single time points (cross-sectional studies) fail to capture the system dynamics and often require larger sample sizes to detect intersample variability. WGS, whole-genome sequencing. Ab, antibody.