Skip to main content

Table 2 Runtime and maximum main memory usage for the construction of the compressed de Bruijn graph

From: A representation of a compressed de Bruijn graph for pan-genome analysis that enables search

k Algorithm 40 E. coli 62 E. coli 7 × Chr1 7 × HG
init SplitMEM 117 (315.25) 141 (317.00)
init A1, A2 38 (5.00) 64 (5.00) 380 (5.00)
init A3, A4 131 (1.32) 202 (1.24) 1168 (1.24) 20,341 (1.24)
50 SplitMEM 2261 (572.19)
50 A1 57 (5.22) 92 (5.34) 596 (6.20)
50 A2 61 (8.49) 97 (8.78) 619 (9.98)
50 A3 188 (2.23) 300 (2.26) 1733 (3.07) 29,816 (2.77)
50 A3compr1 208 (1.81) 346 (1.85) 1880 (2.66) 31,472 (2.36)
50 A3compr2 236 (1.63) 374 (1.66) 2318 (2.51) 39,366 (2.22)
50 A4 164 (1.75) 254 (1.82) 1419 (1.28) 25,574 (1.96)
50 A4compr1 167 (1.46) 257 (1.53) 1435 (1.28) 25,866 (1.66)
50 A4compr2 179 (1.32) 272 (1.24) 1526 (1.24) 27,365 (1.39)
50 A4+explicit 172 (3.26) 268 (3.35) 1515 (3.59) 27,619 (3.88)
50 A4compr1+explicit 176 (2.97) 271 (3.06) 1541 (3.31) 28,044 (3.64)
50 A4compr2+explicit 188 (2.66) 289 (2.74) 1629 (2.96) 29,517 (3.38)
100 SplitMEM 2568 (572.20)
100 A1 59 (5.00) 95 (5.00) 595 (5.95)
100 A2 62 (7.89) 99 (8.19) 605 (9.74)
100 A3 188 (1.63) 299 (1.68) 1738 (2.74) 27,815 (2.23)
100 A3compr1 205 (1.50) 326 (1.49) 1839 (2.33) 30,401 (1.80)
100 A3compr2 232 (1.32) 411 (1.29) 2340 (2.14) 38,134 (1.66)
100 A4 174 (1.71) 261 (1.79) 1422 (1.28) 25,723 (1.94)
100 A4compr1 171 (1.42) 264 (1.50) 1439 (1.28) 26,040 (1.64)
100 A4compr2 185 (1.32) 289 (1.24) 1544 (1.24) 27,464 (1.37)
100 A4+explicit 178 (2.61) 270 (2.73) 1486 (3.21) 26,878 (3.36)
100 A4compr1+explicit 175 (2.32) 273 (2.44) 1500 (2.92) 26,999 (3.07)
100 A4compr2+explicit 190 (2.01) 299 (2.12) 1624 (2.68) 28,665 (2.80)
500 SplitMEM 2116 (570.84)
500 A1 72 (5.00) 113 (5.00) 620 (5.83)
500 A2 83 (7.17) 117 (7.43) 640 (9.66)
500 A3 194 (1.50) 304 (1.49) 1752 (2.67) 28,548 (2.07)
500 A3compr1 216 (1.50) 325 (1.49) 1839 (2.19) 30,488 (1.65)
500 A3compr2 241 (1.32) 378 (1.29) 2319 (2.06) 36,993 (1.50)
500 A4 184 (1.65) 283 (1.74) 1453 (1.28) 26,362 (1.93)
500 A4compr1 197 (1.35) 287 (1.44) 1477 (1.28) 26,545 (1.63)
500 A4compr2 213 (1.32) 322 (1.24) 1622 (1.24) 28,501 (1.36)
500 A4+explicit 185 (1.81) 285 (1.90) 1509 (3.14) 27,285 (3.14)
500 A4compr1+explicit 198 (1.52) 288 (1.61) 1535 (2.83) 27,417 (2.79)
500 A4compr2+explicit 214 (1.32) 323 (1.29) 1694 (2.56) 29,283 (2.58)
  1. The first column shows the k-mer size (an entry init means that only the index data structure is constructed) and the second column specifies the algorithm used in the experiment. The remaining columns show the run-times in seconds and, in parentheses, the maximum main memory usage in bytes per base pair (including the construction) for the data sets described in the text. A minus indicates that the respective algorithm was not able to solve its task on our machine equipped with 128 GB of RAM