Table 4 Performance on real sequencing data
From: Sketching and sampling approaches for fast and accurate long read classification
Total number of reads classified | Number of reads classified to multiple genomes | Number of unclassified reads | Number of reads with same prediction as Minimap2 | |||
|---|---|---|---|---|---|---|
No Threshold | > 50% of read aligned | |||||
Vegan (1.90 M Reads) | Minimap2 (No threshold) | 1,490,713 (78.3%) | 485,233 (25.4%) | 413,446 (21.7%) | N/A | |
Minimap2 (> 50% of read aligned) | 1,069,306 (56.1%) | 23,199 (1.2%) | 834,853 (43.9%) | N/A | ||
Kraken2 | 1,399,341 (73.5%) | 562,744 (29.6%) | 504,818 (26.5%) | 827,572 | 816,331 | |
MinHash | 1,032,396 (54.2%) | 261,732 (13.7%) | 871,763 (45.8%) | 890,766 | 821,556 | |
Minimizer | 1,029,996 (54.1%) | 262,514 (13.8%) | 874,163 (45.9%) | 891,388 | 820,493 | |
Uniform | 1,021,555 (53.6%) | 264,867 (13.9%) | 882,604 (46.7%) | 883,465 | 819,017 | |
Omnivore (1.79 M Reads) | Minimap2 (No threshold) | 1,530,795 (85.4%) | 490,501 (27.4%) | 261,351 (14.6%) | N/A | |
Minimap2 (> 50% of read aligned) | 1,144,452 (63.8%) | 24,271 (1.3%) | 647,694 (36.2%) | N/A | ||
Kraken2 | 1,442,671 (80.5%) | 578,113 (32.2%) | 349,475 (19.5%) | 855,670 | 835,888 | |
MinHash | 1,111,102 (62.0%) | 275,344 (15.4%) | 681,044 (38.0%) | 915,634 | 873,901 | |
Minimizer | 1,105,356 (61.7%) | 278,654 (15.5%) | 686,790 (38.3%) | 916,998 | 872,955 | |
Uniform | 1,098,244 (61.3%) | 281,745 (15.7%) | 693,902 (38.7%) | 911,554 | 871,675 | |