Gene Selection and Classification of scRNA-seq Data Combining Information Gain Ratio and Genetic Algorithm with Dynamic Crossover

Table 5

Purity metric obtained by the IGRDCGA and several competing algorithms.

Dataset	K means	spectral	SIMLR	IGRDCGA	PCA	tSNE
Allodiploid	0.2767	−0.0500	1.0000	1.0000	1.0000	−0.0125
Sasagawa	0.0387	−0.2253	0.5020	0.9170	0.5731	0.1462
Ramskold	0.5953	0.1801	0.7386	0.9756	0.9331	0.5491
Biase	0.5124	0.1796	0.7896	1.0000	0.9636	0.2149
Ting	−0.1852	−0.4401	0.5643	0.4239	0.5204	0.3951
Chung	−0.1686	−0.4338	0.3331	0.1288	0.2860	0.3774
Yeo	−0.0742	−0.4393	0.8247	0.8352	0.8196	0.6856
Ning	−0.1955	−0.3228	0.3667	0.7629	0.3466	0.1128
Ginhoux	0.3983	−0.3019	0.3912	0.6425	0.5308	0.1045
Su	0.4949	−0.4755	0.5004	0.7007	0.5594	0.3739
Usoskin	−0.1365	−0.4286	0.3846	0.8584	0.6343	0.6506
Deng	0.7271	−0.0730	0.6137	0.8920	0.6378	0.2573
Fan	0.1441	−0.2779	0.5980	0.6421	0.6073	0.4231
Goolam	0.1340	−0.1726	0.3632	0.7725	0.2284	0.1841
Kolodz	0.5854	−0.2917	0.9783	1.0000	0.9891	0.1003
Nestorowa	0.5335	−0.0128	0.6124	0.6035	0.6294	0.6444
Treutlein	−0.0020	−0.1692	0.4146	0.8372	0.5156	0.2137
Camp15	0.4989	−0.4715	0.6333	0.6594	0.5855	0.6197
Camp17	0.6779	−0.6129	0.6976	0.7836	0.7586	0.4854
Yan	0.6861	0.3208	0.7748	0.9620	0.9001	0.3921
Wang	−0.3533	−0.5255	0.4835	0.2183	0.4099	0.5145
Li1	0.2909	−0.7039	0.8271	0.8050	0.7802	0.5820
Li2	−0.2315	0.4114	0.4866	0.4227	0.3781	0.4231
Patel	−0.3625	−0.5657	0.7379	0.8244	0.8327	0.8982
Pollen	0.2559	−0.5278	0.9414	0.9271	0.9369	0.9210
Manno_m	0.4832	−0.8098	0.8236	0.7025	0.8149	0.8482
Manno_h	0.6085	0.9307	0.8803	0.8370	0.9099	0.9268
Tasic	0.6178	0.8780	0.8747	0.9231	0.8948	0.8939
Zeisel	0.5849	0.7180	0.7558	0.6948	0.6250	0.7646
Grun	0.2741	0.0728	0.4554	0.6157	0.2331	−0.0003
Baron	0.1026	0.5293	0.5466	0.5121	0.4274	0.5723
Muraro	0.2705	0.7018	0.7430	0.7552	0.4868	0.4388
Xin	0.1638	0.3518	0.2060	0.8166	0.4312	0.1518