Figure 2From: A lightweight, flow-based toolkit for parallel and distributed bioinformatics pipelinesThe produce/spawn/consume idiom. This workflow design pattern, used to process a single input node in parallel, arises in many contexts, such as in replica exchange simulations (see text). In this and remaining workflow diagrams (Figs. 3, 4), the sequence of Piper nodes is shown on the left (A), while the discrete data transformations that will implicitly occur (at the data-item level) are schematized on the right (B).Back to article page