Detailed MATE-CLEVER Pipeline for GoNL

For deletion discovery, we ran the discovery part of MATE-CLEVER [3], with minor modifications that account for volatilities among library protocols. MATE-CLEVER is an integrated approach. Its major purpose in the frame of the project is to discover deletions of size 30–100 bp (sometimes termed the ”twilight zone of NGS indels”). It incorporates CLEVER [2], as an internal segment size based approach that approvedly has state-of-the-art performance rates on indels of size 30–100 bp, and LASER [4], as a split-read aligner. MATE-CLEVER uses CLEVER in a first step, in order to discover deletions of size 30–100 bp at extremely high sensitivity, and uses LASER in a second step, in order to refine the breakpoint annotations made by CLEVER. It also uses several auxiliary tools, as described below. In the following, we provide the full description of details and commands, by which to reproduce MATE-CLEVER’s callset. This pipeline deviates in several details from the description provided in [3] itself, due to the minor modifications mentioned above. To run the CLEVER-based deletion discovery pipeline, revision 3097f2 from the git repository at http://clever-sv.googlecode.com has been used. As above-mentioned, it includes CLEVER [2], LASER [4], and several auxiliary tools described below. To run (and parallelize) the below pipeline, we used the Python-based workflow engine Snakemake [1].