expHRD: an individualized, transcriptome-based prediction model for homologous recombination deficiency assessment in cancer

Table 1 Performance evaluation of various machine-learning methods

	Elastic net		Lasso		Ridge		GBM		SVR (Rbf)		SVR (linear)		MLP
Cancer	34 (33 + TNBC)
Input genes (n)	4436 (result of DEG of OV)
Selected gens (n)	2538		862		4436		2755		x		4436		x
Selected gens (n)	R2	RMSE	R2	RMSE	R2	RMSE	R2	RMSE	R2	RMSE	R2	RMSE	R2	RMSE
PAN	0.7364	9.8624	0.7319	9.9462	0.6375	11.566	0.6943	10.622	0.7238	10.097	0.7163	10.232	0.7129	10.294
OV	0.592	13.125	0.59	13.158	0.53	14.088	0.5408	13.925	0.5568	13.68	0.5235	14.184	0.6243	12.594
TMBC	0.4682	17.765	0.4576	17.941	0.5489	16.361	0.3074	20.273	0.4417	18.201	0.5492	16.355	0.6245	14.927

ISSN: 1471-2105