Pancreas (random split with label noise)

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 randomly with 20% label noise.

Pancreas (random split with label noise)

RankF1 scoreMacro F1 scoreAccuracyMemory (GB)Runtime (min)CPU (%)NamePaperCodeYear
140.170.040.343.604.1813.40Majority Vote

v0.6.0

2022
130.190.070.183.706.1210.00Random Labels

v0.6.0

2022
70.930.820.936.506.08258.60K-neighbors classifier (log CPM)

v1.1.2

1967
10.970.840.976.502.27558.90Logistic regression (log CPM)

v1.1.2

2013
110.880.600.866.507.772,319.90Multilayer perceptron (log CPM)

v1.1.2

1990
20.970.850.978.901.702,098.20XGBoost (log CPM)

v1.6.2

2016
40.970.680.975.8033.851,362.70scANVI (Seurat v3 2000 HVG)

v0.17.1

2021
50.950.600.966.0073.571,307.70scANVI (All genes)

v0.17.1

2021
60.950.750.9522.5020.28150.80K-neighbors classifier (log scran)

v1.1.2

1967
90.940.530.956.6078.081,042.10scArches+scANVI (Seurat v3 2000 HVG)

v0.17.1

2021
120.800.470.8222.5013.50125.40Logistic regression (log scran)

v1.1.2

2013
80.940.530.959.7043.551,473.00scArches+scANVI (All genes)

v0.17.1

2021
30.960.890.9623.1013.97351.00XGBoost (log scran)

v1.6.2

2016
100.930.660.9222.5024.93983.30Multilayer perceptron (log scran)

v1.1.2

1990