Table 3.

Differences in predictive accuracy related to certain versus uncertain sample predictions in the validation set (24 biopsy samples, 7 patients).

Uncertainty threshold	Number of samples (n)	Overall accuracy	Sensitivity/specificity (%/%)
Entire pooled cohort	24	67%	43/76
p < 0.05 (certain)	18	78%	75/79
p > 0.05 (uncertain)	6	33%	0/67

Shown are the differences in predictive performance of the transductive learning GP model for the entire pooled validation cohort and also when stratifying the sample predictions with high certainty (low p values) versus low certainty (high p values) at various p value thresholds.