he multi-statistics multivariate model

ee transposon statistics (the insertion statistic, the site statistic and

tatistic) were joined together, i.e., treated as three variables, to

a multivariate cluster model for gene essentiality analysis for this

The K-means model (kmeans), the fuzzy C-means (cmeans)

Gamma mixture model (mixR) and the Gaussian mixture model

ols) were constructed.

the K-means model and the fuzzy C-means model predicted 442

genes. Four models simultaneously predicted 406 essential genes

these 406 genes had strong essentiality compared with other

gure 2.48 shows the result. In this Venn diagram, it can be seen

disagreement between four models became smaller. The

nt rate between four models was about 86.02% among 472

essential genes by four models.

The Venn diagram of identified essential genes using four multivariate models

hree transposon statistics for the first replicate of the Francisella Tularensis

e 2.49 shows the transposon distribution of an essential gene

18) predicted by four models. The distribution shows that 32

sites were distributed at four sites close to the left distal end.

e, this gene was identified as an essential gene by all four models.

ngle in the inset shows where this gene was in the posterior

ty curve of the mixture Gamma model constructed for the MF

essentiality posterior probability value was almost one.