Workshop Summary: Though many workflow engines exist, it's difficult to port existing descriptions across platforms or scale, reproduce, or recreate data analyses. The Common Workflow Language (CWL) consists of various organizations and individuals that have an interest in portability and reproducibility of data analysis workflows. The goal is to create a specification that enable data scientists to create analyses that are powerful, easy to use, portable, and support reproducibility. CWL builds on technologies such as JSON-LD, Avro, and Docker and is designed to express workflows for data-intensive fields, such as Bioinformatics, Chemistry, Physics, and Astronomy. We will discuss the Seven Bridges Cancer Genomics Cloud (CGC) as a CWL case study. Each of the tools and pipelines used to analyzes more than a petabyte of data on the CGC are built in CWL and are portable to conformant platforms.
The full workshop schedule can be found here.