Skip to main content
. 2022 Feb 15;11:giac002. doi: 10.1093/gigascience/giac002

Table 3:

Runtimes for DDN generation given input datasets with different numbers of phenotypes

Server runtime to generate network after receiving HTTP request (sec)
Phenotype count Fruchterman-Reingold layout Force Atlas 2 layout
1 2 3 4 5 Mean (SD) 1 2 3 4 5 Mean (SD)
50 3.07 2.34 2.86 2.31 2.76 2.67 (0.33) 2.46 2.48 2.93 2.43 3.00 2.66 (0.28)
100 3.26 3.49 4.29 3.61 3.52 3.63 (0.39) 3.43 4.14 4.37 4.62 3.58 4.03 (0.51)
250 6.60 5.20 6.77 6.62 5.56 6.15 (0.72) 6.74 5.31 6.36 6.92 5.90 6.25 (0.65)
500 11.21 11.85 12.53 10.94 9.91 11.29 (0.99) 11.68 12.04 12.49 11.21 9.33 11.35 (1.22)
1,000 28.27 28.77 30.19 27.01 29.52 28.75 (1.22) 29.37 28.35 29.84 27.23 30.23 29.00 (1.22)
UKBB DDN 48.60 N/A 39.43 N/A

These times measure how long it takes for the server to generate the network after the “submit” button has been clicked—in all instances, files have already been uploaded to the server. Upload speeds for files will vary depending on user bandwidth. Five different datasets were constructed for each count of phenotypes to evaluate runtime, and the mean and standard deviation of time for the 5 runs is also provided for each row. Finally, runtime for the full input UKBB case study is included in the last row of the table. N/A: not applicable.