Skip to main content

Table 4 Summary of internal evaluations for each compared methods

From: A comparison framework and guideline of clustering methods for mass cytometry data

DatasetsMethodsInternal evaluations
CHDBXB
Cell CycleAccense3.2409 ± 0.20692.4746 ± 0.25630.6473 ± 0.0750
PhenoGraph3.4250 ± 0.03502.5692 ± 0.15030.7864 ± 0.0392
Xshift3.5620 ± 0.04431.7337 ± 0.16180.9131 ± 0.3066
kmeans3.9414 ± 0.00111.5971 ± 0.00130.7705 ± 0.0165
flowMeans3.5992 ± 0.02831.6237 ± 0.03410.7294 ± 0.0393
FlowSOM3.6294 ± 0.00361.3578 ± 0.03510.6669 ± 0.0269
DEPECHE3.8372 ± 0.02831.7767 ± 0.03220.7695 ± 0.0177
ACDC3.4645 ± 0.09012.0773 ± 0.31560.9395 ± 0.0428
LDA3.4581 ± 0.00362.5080 ± 0.01911.0021 ± 0.0358
ColonAccense3.2688 ± 0.06221.7064 ± 0.06561.0427 ± 0.0545
PhenoGraph3.4231 ± 0.01941.9030 ± 0.02451.3997 ± 0.1199
Xshift3.0706 ± 0.01792.3218 ± 0.05951.3207 ± 0.0511
kmeans3.5109 ± 0.00791.9037 ± 0.05011.4570 ± 0.1055
flowMeans3.4793 ± 0.10681.5864 ± 0.04211.6706 ± 0.1958
FlowSOM3.6435 ± 0.01171.6583 ± 0.05141.5910 ± 0.1568
DEPECHE3.9819 ± 0.03751.7725 ± 0.02501.6325 ± 0.0930
ACDC3.5898 ± 0.04402.0607 ± 0.11601.4239 ± 0.0754
LDA3.5100 ± 0.00092.1773 ± 0.00651.5564 ± 0.0724
Levine13dimAccense3.4230 ± 0.09321.8832 ± 0.14081.2321 ± 0.0278
PhenoGraph4.0739 ± 0.01761.4645 ± 0.03461.3972 ± 0.1349
Xshift3.5106 ± 0.02892.4284 ± 0.04431.7868 ± 0.0476
kmeans3.8508 ± 0.01502.1550 ± 0.05461.6213 ± 0.1471
flowMeans4.0475 ± 0.01941.5030 ± 0.08491.4234 ± 0.1182
FlowSOM3.8486 ± 0.00711.7564 ± 0.06151.5043 ± 0.1531
DEPECHE4.2783 ± 0.01741.1677 ± 0.03421.3562 ± 0.0392
ACDC3.9638 ± 0.01101.4916 ± 0.03701.3109 ± 0.0948
LDA3.8288 ± 0.01062.0046 ± 0.04931.3828 ± 0.1167
Levine32dimAccense3.4621 ± 0.09012.3414 ± 0.09250.7891 ± 0.0950
PhenoGraph3.7401 ± 0.00811.8293 ± 0.08101.0009 ± 0.0479
Xshift3.6669 ± 0.01022.2576 ± 0.13240.8295 ± 0.1382
kmeans3.8761 ± 0.01662.0587 ± 0.03860.9972 ± 0.0441
flowMeans3.8546 ± 0.03931.6975 ± 0.21990.7985 ± 0.0709
FlowSOM3.8244 ± 0.02851.5974 ± 0.08630.8366 ± 0.0792
DEPECHE4.1480 ± 0.00091.4727 ± 0.00230.7575 ± 0.0351
ACDC3.6169 ± 0.00461.3974 ± 0.00490.7693 ± 0.1310
LDA3.8297 ± 0.00071.7011 ± 0.00990.7155 ± 0.0139
MuscleAccense3.2254 ± 0.16882.3190 ± 0.31780.8420 ± 0.1211
PhenoGraph3.6052 ± 0.01321.8619 ± 0.04171.7228 ± 0.2389
Xshift3.2898 ± 0.02742.2460 ± 0.15242.1455 ± 0.1599
kmeans3.9722 ± 0.00221.7729 ± 0.00991.4987 ± 0.1212
flowMeans3.3809 ± 0.05651.7685 ± 0.07691.3750 ± 0.1428
FlowSOM3.8262 ± 0.01461.4439 ± 0.04621.3586 ± 0.1473
DEPECHE4.2639 ± 0.00441.2235 ± 0.00801.2892 ± 0.0269
ACDC3.6900 ± 0.03051.7186 ± 0.05151.5641 ± 0.15555
LDA3.8073 ± 0.00311.6641 ± 0.02061.6003 ± 0.0900
Samusik01Accense3.2952 ± 0.04762.3424 ± 0.14540.6500 ± 0.0266
PhenoGraph3.7380 ± 0.00331.4607 ± 0.04240.7669 ± 0.0831
Xshift3.2660 ± 0.01342.7855 ± 0.46441.2670 ± 0.1654
kmeans3.7087 ± 0.01352.2072 ± 0.08540.9365 ± 0.0792
flowMeans3.4029 ± 0.03811.6069 ± 0.23740.6309 ± 0.0475
FlowSOM3.6941 ± 0.02571.8967 ± 0.37990.8286 ± 0.0832
DEPECHE4.1028 ± 0.00211.4867 ± 0.00540.9141 ± 0.0109
ACDC3.6827 ± 0.01781.3871 ± 0.01320.7273 ± 0.0573
LDA3.6767 ± 0.00801.6325 ± 0.02421.0414 ± 0.0808
  1. Data shown as mean ± standard deviation. CH Calinski-Harabasz index (log10 transformed), DB Davies-Bouldin index, XB Xie-Beni index (log10 transformed)