Table 3.
Identification of never-smokers based on supervised analysis of never-smokers versus smokers in seven AC data sets
| DCCA | GSE10072A | GSE11969A | GSE12667A | GSE32863A | BeerA | IlluminaA* | |
|---|---|---|---|---|---|---|---|
| DCC centroids |
88%** |
88% |
93% |
100% |
86% |
89% |
80% |
|
GSE10072 centroids |
90% |
100%** |
87% |
100% |
76% |
100% |
80% |
|
GSE11969 centroids |
78% |
88% |
89%** |
100% |
59% |
89% |
80% |
|
GSE12667 centroids |
76% |
88% |
82% |
100%** |
66% |
56% |
60% |
|
GSE32863 centroids |
78% |
81% |
93% |
88% |
97%** |
100% |
90% |
| Beer centroids |
84% |
94% |
87% |
100% |
76% |
100%** |
70% |
| Illumina centroids * | 76% | 75% | 89% | 88% | 76% | 78% | 100%** |
A Sensitivity in identifying true never-smokers of a supervised classifier (rows) when applied to a specific data set (columns).
* The original cohort of 39 AC.
** Classifier applied to its own training data.