Skip to main content
. 2020 Jul 15;97(11):1180–1186. doi: 10.1002/cyto.a.24187

Figure 1.

Figure 1

A. In total, 211,359 publicly available FCS files were downloaded (RAW‐FCS), consisting of 206,620 single data set files (RAW‐SD‐FCS) and 4,739 multi data set files (RAW‐MD‐FCS). Each RAW‐FCS was checked for conformance with the FCS specification. For each RAW‐FCS, multiple artificial FCS files were created (details in B and C), allowing in depth evaluation of various aspects of RAW‐FCS, in a highly space efficient manner. For each data set, within each RAW‐FCS, validated parse results were obtained for the TEXT segment, by aggregating FlowCore and FlowJo parse results. The metadata of 211,359 public FCS files can now be easily explored. Together, the validated parse results and artificial FCS files form a fully fledged testing set for FCS parser development. Performance of FlowCore was benchmarked in‐depth. B. SLICED‐FCS mimic RAW‐SD‐FCS, as they are identical to RAW‐SD‐FCS, except for the bytes after the TEXT segment being removed. INIT‐FCS are newly created FCS files, which contain the initial TEXT segment from RAW‐FCS, preceded by a HEADER with accurate offsets. TEST‐FCS contain the first 20 events of the original DATA segment, and a slightly modified version of the original TEXT segment (reflecting the new DATA segment size). C. BLANKED‐FCS mimic RAW‐MD‐FCS, as they are identical to RAW‐MD‐FCS, except for the bytes from the DATA and ANALYSIS segment being replaced by 0x20 bytes (space character in ASCII). ADD‐FCS are similar to INIT‐FCS, but contain the additional TEXT segment(s) from RAW‐MD‐FCS.