Combined performance of screening and variable selection methods in ultra-high dimensional data in predicting time-to-event outcomes

Table 7 c-indices based on the training and testing sets for the real example

Selection approach	No of SNPs selected	Training set (419)		Testing set (n = 204)
Selection approach	No of SNPs selected	Original c-index	Corrected c-index*	c-index* (95% CI)
ISIS-LASSO	2	0.649	0.646	0.664 (0.621–0.707)
ISIS-ALASSO	0	0.649	0.650	0.671 (0.624–0.719)
ISIS-RSF	2	0.650	0.645	0.669 (0.618–0.720)
SIS	2	0.650	0.646	0.669 (0.624–0.714)
SIS-LASSO	0	0.649	0.649	0.671 (0.626–0.717)
SIS-ALASSO	0	0.649	0.650	0.671 (0.620–0.723)
SIS-RSF	2	0.650	0.648	0.669 (0.623–0.716)
PSIS	40	0.749	–	0.572 (0.527–0.617)
PSIS-LASSO	28	0.746	0.727	0.568 (0.524–0.613)
PSIS-ALASSO	24	0.744	0.727	0.573 (0.528–0.617)
PSIS-RSF	35	0.748	–	0.575 (0.529–0.622)
LASSO	16	0.716	–	0.586 (0.540–0.632)
ALASSO	13	0.653	0.634	0.647 (0.601–0.693)

ISSN: 2397-7523