Table 1 Open in new tab Classification...

Table 1

Classification tools selected for evaluation in this article and their information

Tools		(Pre)published date	Feature selection	Algorithm	Data format	‘Unassigned’ function
Scmap	scmapc2c	02/04/2018	Yes#	k-Means and approximate KNN, cosine distance	Normalized counts/log counts	Yes
Scmap	scmapc2clus	02/04/2018	Yes#	cluster-level median expression, cosine distance	Normalized counts/log counts	Yes
scMCA		02/22/2018	No	Cluster-level mean expression, Pearson correlation	Log counts	No
scPred		07/14/2018*	Yes	SVM	Normalized counts → cpm	Yes
SingleR		01/14/2019	Yes#	Cluster-level median expression, Spearman correlation	(Normalized) counts	No
Seurat		04/13/2015	Yes	RandomForest	(Normalized) counts	No
CaSTLe		10/10/2018	Yes	XGBoost	Log counts	No
scID		11/14/2018*	Yes#	A two-mixed Gaussian distribution	Counts → norm counts (with build-in function)	Yes
AltAnalyze		08/31/2016	Yes#	No description for algorithm	Norm counts	No
CellFishing		11/29/2018	Yes#	approximate k-NN, Locality- sensitive hashing and Hamming distance	Raw counts	No

Tools		(Pre)published date	Feature selection	Algorithm	Data format	‘Unassigned’ function
Scmap	scmapc2c	02/04/2018	Yes#	k-Means and approximate KNN, cosine distance	Normalized counts/log counts	Yes
Scmap	scmapc2clus	02/04/2018	Yes#	cluster-level median expression, cosine distance	Normalized counts/log counts	Yes
scMCA		02/22/2018	No	Cluster-level mean expression, Pearson correlation	Log counts	No
scPred		07/14/2018*	Yes	SVM	Normalized counts → cpm	Yes
SingleR		01/14/2019	Yes#	Cluster-level median expression, Spearman correlation	(Normalized) counts	No
Seurat		04/13/2015	Yes	RandomForest	(Normalized) counts	No
CaSTLe		10/10/2018	Yes	XGBoost	Log counts	No
scID		11/14/2018*	Yes#	A two-mixed Gaussian distribution	Counts → norm counts (with build-in function)	Yes
AltAnalyze		08/31/2016	Yes#	No description for algorithm	Norm counts	No
CellFishing		11/29/2018	Yes#	approximate k-NN, Locality- sensitive hashing and Hamming distance	Raw counts	No

Note. Dates labeled with ‘*’ mean the preprint date of the corresponding tool. The # label means corresponding tools have the option to perform the feature selection using a user-defined gene list. The published date of Seurat and AltAnalyze are the published dates for the packages but not for their classification functions. The version of all tools adopted in this article was up to date as of 31 December 2018.

This Feature Is Available To Subscribers Only