Pancreas (by batch)

Human pancreatic islet scRNA-seq data from 6 datasets across technologies (CEL-seq, CEL-seq2, Smart-seq2, inDrop, Fluidigm C1, and SMARTER-seq). Split into train/test by experimental batch. Dimensions: 16382 cells, 18771 genes. 14 cell types (avg. 1170±1703 cells per cell type).

Raw Scaled

Rank	Name	Mean score	Accuracy	Accuracy	F1 score	F1 score	Macro F1 score	Macro F1 score	Memory (GB)	Runtime (min)	CPU (%)	Year	Library
1	Multilayer perceptron (log CP10k)	0.95	0.96	0.95	0.96	0.95	0.94	0.93	2.30	3.05	704.80	1990	v1.1.3
2	Logistic regression (log CP10k)	0.92	0.97	0.96	0.97	0.96	0.86	0.85	2.30	4.28	374.30	2013	v1.1.3
3	XGBoost (log CP10k)	0.91	0.94	0.92	0.94	0.92	0.91	0.90	2.70	11.87	294.10	2016	v1.6.2
4	Seurat reference mapping (SCTransform)	0.91	0.96	0.95	0.96	0.95	0.85	0.84	445.50	13.77	125.40	2021	v4.3.0
5	Multilayer perceptron (log scran)	0.87	0.95	0.94	0.95	0.94	0.76	0.74	8.30	12.12	936.00	1990	v1.1.3
6	XGBoost (log scran)	0.87	0.93	0.92	0.93	0.92	0.79	0.77	8.50	12.03	328.40	2016	v1.6.2
7	scANVI (All genes)	0.84	0.96	0.95	0.96	0.94	0.66	0.64	4.50	43.48	2,440.00	2021	v0.20.0
8	scANVI (Seurat v3 2000 HVG)	0.84	0.96	0.95	0.96	0.95	0.66	0.63	6.90	32.87	2,380.40	2021	v0.20.0
9	scArches+scANVI (All genes)	0.83	0.95	0.94	0.95	0.94	0.65	0.62	5.80	24.17	2,612.90	2021	v0.20.0
10	K-neighbors classifier (log CP10k)	0.82	0.88	0.84	0.88	0.84	0.79	0.77	2.30	2.75	298.10	1967	v1.1.3
11	Logistic regression (log scran)	0.81	0.94	0.92	0.94	0.92	0.63	0.60	8.50	16.78	131.20	2013	v1.1.3
12	scArches+scANVI (Seurat v3 2000 HVG)	0.78	0.95	0.94	0.94	0.93	0.52	0.48	4.00	21.87	1,358.50	2021	v0.20.0
13	K-neighbors classifier (log scran)	0.72	0.84	0.80	0.84	0.80	0.58	0.55	8.30	8.77	118.70	1967	v1.1.3
14	Majority Vote	0.04	0.35	0.19	0.18	-0.03	0.04	-0.03	0.69	1.08	19.90	2022	v0.7.0