Table 9:

Summary of the performance improvements obtained by parSMURF by tuning its learning parameters with the GWAS Catalog dataset

Parameters	AUROC mean (SD)	AUPRC mean (SD)
Default parameters	0.99426 (0.00169)	0.48058 (0.07138)
Grid search	0.99459 (0.00174)	0.72533 (0.03616)
Bayesian optimizer	0.99346 (0.00193)	0.71945 (0.03675)

Results are averaged across 10 repetitions of the 10-fold cytoband-aware cross-validation. AUROC and AUPRC are averaged across the 10 folds. Default parameters: nParts 100, fp 2, ratio 3, nTrees 10, mTry 5.