Table 7

Observed output variability

50 × 60 seqs mut stacked

50 × 60 seqs mut interleaved

Prairie soil

N

μ

Err

Min

Max

μ

Err

Min

Max

μ

Err

Min

Max


CROP 3%

20

1852.2

2.16

1847.6

1856.7

1954.6

0.83

1952.8

1956.3

1128

0

1128

1128

CROP 5%

20

1853.5

2.43

1848.4

1858.5

1952.4

0.66

1951

1953.7

1128

0

1128

1128

CROP 10%

20

1850.9

1.65

1847.4

1854.4

1951.7

0.62

1950.4

1953

1128

0

1128

1128

ESPRIT 3%

20

205

0

205

205

193

0

193

193

565

0

565

565

ESPRIT 5%

20

59

0

59

59

59

0

59

59

381

0

381

381

ESPRIT 10%

20

56

0

56

56

56

0

56

56

180

0

180

180

Mothur 3%

20

1465.3

0.23

1464.8

1465.8

1173

0

1173

1173

541

0

541

541

Mothur 5%

20

1465.3

0.23

1464.8

1465.8

1173

0

1173

1173

400

0

400

400

Mothur 10%

20

1465.3

0.23

1464.8

1465.8

1173

0

1173

1173

347

0

347

347

Mothur+PreC 3%

20

1490.6

0.16

1490.3

1490.9

1197.3

0.12

1197

1197.6

541

0

541

541

Mothur+PreC 5%

20

1490.6

0.16

1490.3

1490.9

1197.3

0.12

1197

1197.6

400

0

400

400

Mothur+PreC 10%

20

1490.6

0.16

1490.3

1490.9

1197.3

0.12

1197

1197.6

348

0

348

348


Observed output variability with 20 equal runs of CROP, ESPRIT and Mothur (with (PreC) and without preclustering). N is the number of observations for all samples; μ is the observed mean of the sample. Standard error (Err) and confidence intervals (Min and Max) were calculated from these values, as described in Methods. Many of the CROP calculations failed to complete in this test, as reflected by N (Prairie soil), and statistics were adjusted accordingly.

Barriuso et al. BMC Bioinformatics 2011 12:473   doi:10.1186/1471-2105-12-473

Open Data