Mash Screen: high-throughput sequence containment estimation for genome discovery

Table 3 Potential reduction in classification database size by pre-screening, illustrated using constituents of the Shakya metagenome

Containment (≥)	Positives	Positives (%)	Database genomes	Database gigabases
0	58/58	100	83,327	330
0.9	58/58	100	1585	5.76
0.99	56/58	96	227	0.878
0.999	53/58	91	76	0.237
0.9999	43/58	74	58	0.182

Containment refers to the threshold of the Mash containment score used to filter the genomes in the database, with 0 meaning no filtering and 0.9 meaning filter all genomes with a containment of less than 90%. Positives refers to how many of the known constituents had scores above that threshold, and thus would pass pre-screening (6 of the 64 constituents were not included because they have since been removed from RefSeq and thus were not in the reference database). Database genomes refers to the total number of genomes in the database with scores above the threshold, and Database gigabases refers to the total number of bases in those genomes

ISSN: 1474-760X