optimalFlow: optimal-transport approach to Flow Cytometry analysis
01 November, 2022
Contents
1 Introduction
optimalFlow is a package dedicated to applying optimal-transport techniques to supervised flow cytometry gating based on the results in del Barrio et al. (2019).
We provide novel methods for grouping (clustering) gated cytometries. By clustering a set of cytometries we are producing groups (clusters) of cytometries that have lower variability than the whole collection. This in turn allows to improve greatly the performance of any supervised learning procedure. Once we have a partition (clustering) of a collection of cytometries, we provide several methods for obtaining an artificial cytometry (prototype, template) that represents in some optimal way the cytometries in each respective group. These prototypes can be used, among other things, for matching populations between different cytometries. Even more, a procedure able to group similar cytometries could help to detect individuals with a common particular condition, for instance some kind of disease.
optimalFlowTemplates is our procedure for clustering cytometries and obtaining templates. It is based on recent developments in the field of optimal transport such as a similarity distance between clusterings and a barycenter (Frechet mean) and k-barycenters of probability distributions.
We introduce optimalFlowClassification, a supervised classification tool for the case when a database of gated cytometries is available. The procedure uses the prototypes obtained by optimalFlowTemplates on the database. These are used to initialize tclust, a robust extension of k-means that allows for non-spherical shapes, for gating a new cytometry (see Garcia-Escudero et al. (2008)). By using a similarity distance between the best clustering obtained by tclust and the artificial cytometries provided by optimalFlowTemplates we can assign the new cytometry to the most similar template (and the respective group of cytometries). We provide several options of how to assign cell types to the new cytometry using the most relevant information, represented by the assigned template and the respective cluster of cytometries.
2 Installation
Installation procedure:
if (!requireNamespace("BiocManager", quietly = TRUE))
install.packages("BiocManager")
BiocManager::install("optimalFLow")3 optimalFlowTemplates
library(optimalFlowData)
library(optimalFlow)
library(ellipse)We start by providing a database of gated cytometries. In this case we select as a learning set 15 cytometries of healthy individuals, from the data provided in optimalFlowData. We will use Cytometry1 to test the results of our procedures. For simplicity and for the sake of a good visualisation we will select only some of the cell types, in particular a subset of 4 cell types.
database <- buildDatabase(
dataset_names = paste0('Cytometry', c(2:5, 7:9, 12:17, 19, 21)),
population_ids = c('Monocytes', 'CD4+CD8-', 'Mature SIg Kappa', 'TCRgd-'))Then we apply optimalFlowTemplates to obtain a clustering of the database and a template cytometry for each group.
templates.optimalFlow <-
optimalFlowTemplates(
database = database
)When running the default mode for optimalFlowTemplates we obtain a plot as in the figure bellow and then we are asked how many clusters we want to look for.
From the plot it seems reasonable to look for 5 clusters of cytometries and we could introduce 5 and press enter, and the procedure will give us a clustering of the learning database and the respective templates. Since this is hard to show in a vignette, an equivalent way of doing this procedure is to execute the command bellow, where we ask for 5 clusters directly.
templates.optimalFlow <-
optimalFlowTemplates(
database = database, templates.number = 5, cl.paral = 1
)## [1] "step 1: 10.0325977802277 secs"
## [1] "step 2: 4.66525173187256 secs"
## [1] "Execution time: 14.6988894939423 secs"Now let us understand what does optimalFlowTemplates return. In the entry templates we have the artificial cytometries, viewed as mixtures of multivariate normal distributions, corresponding to the clustering of the cytometries in the database argument.
length(templates.optimalFlow$templates) # The number of clusters, and, hence, of templates ## [1] 5length(templates.optimalFlow$templates[[1]]) # The number of elements of the first template, it contains four cell types## [1] 4templates.optimalFlow$templates[[1]][[1]] # The first element of the first template## $mean
## [1] 2192.494 3810.563 6952.128 5639.010 5384.128 2326.237 5922.809 2433.990
## [9] 1616.252 1213.455
##
## $cov
## [,1] [,2] [,3] [,4] [,5] [,6]
## [1,] 266639.3056 -22292.14 -239.0103 535.9488 1567.078 6667.185
## [2,] -22292.1404 836892.17 139258.7106 -11083.0742 -26847.427 -85912.718
## [3,] -239.0103 139258.71 84833.6993 2831.9105 -1760.813 -19372.852
## [4,] 535.9488 -11083.07 2831.9105 16117.5767 9840.943 9733.885
## [5,] 1567.0784 -26847.43 -1760.8132 9840.9435 17040.124 19173.992
## [6,] 6667.1845 -85912.72 -19372.8520 9733.8847 19173.992 260893.832
## [7,] -10530.5174 -12155.85 -12970.2311 8010.5740 11112.548 24739.592
## [8,] 6396.7471 -53880.87 -11968.9063 7805.0808 10494.817 63602.554
## [9,] 3686.7700 -83687.89 -17876.9602 13542.7223 17424.684 43574.279
## [10,] 2391.6169 -61685.79 -13296.8697 13933.2554 17777.542 40569.455
## [,7] [,8] [,9] [,10]
## [1,] -10530.517 6396.747 3686.77 2391.617
## [2,] -12155.854 -53880.868 -83687.89 -61685.794
## [3,] -12970.231 -11968.906 -17876.96 -13296.870
## [4,] 8010.574 7805.081 13542.72 13933.255
## [5,] 11112.548 10494.817 17424.68 17777.542
## [6,] 24739.592 63602.554 43574.28 40569.455
## [7,] 135789.172 7525.035 21765.08 20272.825
## [8,] 7525.035 153027.526 25157.58 26236.325
## [9,] 21765.079 25157.581 76070.14 43082.883
## [10,] 20272.825 26236.325 43082.88 51273.104
##
## $weight
## [1] 0.5845872
##
## $type
## [1] "CD4+CD8-"In the argument clustering we have the clustering of the cytometries in the database argument.
templates.optimalFlow$clustering## [1] 1 2 3 3 3 3 4 4 4 1 2 3 5 5 5In the argument database.elliptical we have a list containing each cytometry in the database viewed as a mixture distribution. Each element of the list is a cytometry viewed as a mixture.
length(templates.optimalFlow$database.elliptical) # the number of elements in the database## [1] 15length(templates.optimalFlow$database.elliptical[[1]]) # the number of cell types in the first element of the database## [1] 4templates.optimalFlow$database.elliptical[[1]][[1]] # the parameters corresponding to the first cell type in the first cytometry of the database ## $mean
## CD19/TCRgd:PE Cy7-A LOGICAL CD38:APC H7-A LOGICAL
## 2143.244 3769.941
## CD3:APC-A LOGICAL CD4+CD20:PB-A LOGICAL
## 6912.857 5613.595
## CD45:PO-A LOGICAL CD56+IgK:PE-A LOGICAL
## 5404.314 2068.134
## CD5:PerCP Cy5-5-A LOGICAL CD8+IgL:FITC-A LOGICAL
## 5784.786 2403.888
## FSC-A LINEAR SSC-A Exp-SSC Low
## 1578.943 1226.318
##
## $cov
## CD19/TCRgd:PE Cy7-A LOGICAL CD38:APC H7-A LOGICAL
## CD19/TCRgd:PE Cy7-A LOGICAL 304347.8532 -31430.672
## CD38:APC H7-A LOGICAL -31430.6717 786050.701
## CD3:APC-A LOGICAL -2056.8937 150481.538
## CD4+CD20:PB-A LOGICAL 666.2252 -10919.205
## CD45:PO-A LOGICAL 2933.2678 -34008.120
## CD56+IgK:PE-A LOGICAL 7953.3654 -68622.841
## CD5:PerCP Cy5-5-A LOGICAL -7811.8852 5441.639
## CD8+IgL:FITC-A LOGICAL 4737.6958 -42476.068
## FSC-A LINEAR 3825.7518 -87894.919
## SSC-A Exp-SSC Low 3422.1236 -72960.311
## CD3:APC-A LOGICAL CD4+CD20:PB-A LOGICAL
## CD19/TCRgd:PE Cy7-A LOGICAL -2056.894 666.2252
## CD38:APC H7-A LOGICAL 150481.538 -10919.2047
## CD3:APC-A LOGICAL 93068.194 2462.9102
## CD4+CD20:PB-A LOGICAL 2462.910 15537.7342
## CD45:PO-A LOGICAL -4062.620 9567.5234
## CD56+IgK:PE-A LOGICAL -11052.254 8858.6946
## CD5:PerCP Cy5-5-A LOGICAL -6831.220 8039.8935
## CD8+IgL:FITC-A LOGICAL -8281.683 7577.5080
## FSC-A LINEAR -21798.893 14235.2106
## SSC-A Exp-SSC Low -17098.563 15356.9021
## CD45:PO-A LOGICAL CD56+IgK:PE-A LOGICAL
## CD19/TCRgd:PE Cy7-A LOGICAL 2933.268 7953.365
## CD38:APC H7-A LOGICAL -34008.120 -68622.841
## CD3:APC-A LOGICAL -4062.620 -11052.254
## CD4+CD20:PB-A LOGICAL 9567.523 8858.695
## CD45:PO-A LOGICAL 18350.645 23190.956
## CD56+IgK:PE-A LOGICAL 23190.956 305557.028
## CD5:PerCP Cy5-5-A LOGICAL 10778.661 25521.650
## CD8+IgL:FITC-A LOGICAL 10729.306 36294.993
## FSC-A LINEAR 20416.550 42905.671
## SSC-A Exp-SSC Low 22260.189 46700.463
## CD5:PerCP Cy5-5-A LOGICAL CD8+IgL:FITC-A LOGICAL
## CD19/TCRgd:PE Cy7-A LOGICAL -7811.885 4737.696
## CD38:APC H7-A LOGICAL 5441.639 -42476.068
## CD3:APC-A LOGICAL -6831.220 -8281.683
## CD4+CD20:PB-A LOGICAL 8039.894 7577.508
## CD45:PO-A LOGICAL 10778.661 10729.306
## CD56+IgK:PE-A LOGICAL 25521.650 36294.993
## CD5:PerCP Cy5-5-A LOGICAL 132092.612 2550.771
## CD8+IgL:FITC-A LOGICAL 2550.771 152836.556
## FSC-A LINEAR 22131.202 20286.385
## SSC-A Exp-SSC Low 21582.260 25351.863
## FSC-A LINEAR SSC-A Exp-SSC Low
## CD19/TCRgd:PE Cy7-A LOGICAL 3825.752 3422.124
## CD38:APC H7-A LOGICAL -87894.919 -72960.311
## CD3:APC-A LOGICAL -21798.893 -17098.563
## CD4+CD20:PB-A LOGICAL 14235.211 15356.902
## CD45:PO-A LOGICAL 20416.550 22260.189
## CD56+IgK:PE-A LOGICAL 42905.671 46700.463
## CD5:PerCP Cy5-5-A LOGICAL 22131.202 21582.260
## CD8+IgL:FITC-A LOGICAL 20286.385 25351.863
## FSC-A LINEAR 73060.828 50515.337
## SSC-A Exp-SSC Low 50515.337 59919.397
##
## $weight
## [1] 0.6898461
##
## $type
## [1] "CD4+CD8-"In order to get some intuition about our methodology we are going to give some visual examples. Users can do it for their own data once they have applied optimalFlowTemplates.
We start with a two-dimensional representation of the cytometries of the cluster labelled as 3. As we have gated cytometries in the database we know every cell type, and, even more, we can consider every cytometry as a mixture of multivariate Gaussian distributions and this is stored in templates.optimalFlow$database.elliptical. The user just has to select the variables in which to project the cytometries through the variable dimensions.
cytoPlotDatabase(templates.optimalFlow$database.elliptical[which(templates.optimalFlow$clustering == 3)], dimensions = c(4,3), xlim = c(0, 8000), ylim = c(0, 8000), xlab = "", ylab = "")
Black ellipses correspond to the cell type CD4+CD8- in each cytometry and enclose 95% of the probability for the respective multivariate normal distributions. Red ellipses correspond to Mature Sig Kappa and so on.
A three-dimensional plot of the same case is provided as a static image and can be obtained using the following code.
cytoPlotDatabase3d(templates.optimalFlow$database.elliptical[which(templates.optimalFlow$clustering == 3)], dimensions = c(4, 3, 9), xlim = c(0, 8000), ylim = c(0, 8000), zlim = c(0, 8000))
Figure 1: pooling database
optimalFlowTemplates provides a template cytometry for each cluster, stored in the entry templates. We present here how to visualize in 2d the consensus cytometry, the template, corresponding to cluster 3. Recall that the cytometries belonging to cluster 3 have been plotted above. The code is straightforward, we access templates in templates.optimalFlow and select the third element of the list, since we are interested in cluster 3.
cytoPlot(templates.optimalFlow$templates[[3]], dimensions = c(4,3), xlim = c(0, 8000), ylim = c(0, 8000), xlab = "", ylab = "")
A three dimensional plot of the same case is provided as a static image and can be obtained using the following code.
cytoPlot3d(templates.optimalFlow$templates[[3]], dimensions = c(4, 3, 9), xlim = c(0, 8000), ylim = c(0, 8000), zlim = c(0, 8000))
cosa
It is clear that the prototype cytometry represents well the geometric information of the respective group of cytometries. This visualisation schemes allow users to check by themselves if the templates that they are obtaining are satisfying and if their clusters are really homogenous.
Another relevant situation in flow cytometry is when gatings of cytometries are available but without the identification of each cell type. For the cytometries of cluster 3 it is like we forgot about the colour, since now we do not have cell types identified.
cytoPlotDatabase(templates.optimalFlow$database.elliptical[which(templates.optimalFlow$clustering == 3)], dimensions = c(4,3), xlim = c(0, 8000), ylim = c(0, 8000), xlab = "", ylab = "", colour = FALSE)
The respective 3d static image can be obtained using the following code.
cytoPlotDatabase3d(templates.optimalFlow$database.elliptical[which(templates.optimalFlow$clustering == 3)], dimensions = c(4, 3, 9), xlim = c(0, 8000), ylim = c(0, 8000), zlim = c(0, 8000), colour = FALSE)
From a visual inspection there is enough geometrical information that allows us to differentiate the cel types. It is just a matter of how to capture it.
Indeed, using some unsupervised procedure to obtain the consensus element (the prototype cytometry) should be enough to capture the relevant cluster structure. This can be achieved using otpimalFlowTemplates as follows.
In the following chunk of code we are using optimalFlowTemplates on our database, we are again looking for 5 clusters with the default clustering procedure which is hierarchical complete-linkage but now we vary the consensus.method variable. We are selecting to obtain the template cytometry using k-barycenters in the Wasserstein space, where k is set to be 4.
templates.optimalFlow.barycenter <-
optimalFlowTemplates(
database = database, templates.number = 5, consensus.method = "k-barycenter",
barycenters.number = 4, bar.repetitions = 10, alpha.bar = 0.05, cl.paral = 1
)## [1] "step 1: 12.4316513538361 secs"
## [1] "step 2: 2.16795048713684 mins"
## [1] "Execution time: 2.37515055735906 mins"A different way of obtaining the consensus cytometry is to use density based hierarchical clustering, in this case hdbscan, setting consensus.method = “hierarchical”. The advantage of this is that the number of cell types in the template cytometry is selected automatically.
templates.optimalFlow.hdbscan <-
optimalFlowTemplates(
database = database, templates.number = 5, consensus.method = "hierarchical",
cl.paral = 1
)## [1] "step 1: 17.9575810432434 secs"
## [1] "step 2: 2.17344868977865 mins"
## [1] "Execution time: 2.47275014718374 mins"As before, we can check how well the prototypes represent the group of cytometries. Again, we work with cluster 3. A 2d plot of the prototype cytometry obtained when using a 4-barycenter is provided, where colours represent different groups.
cytoPlot(templates.optimalFlow.barycenter$templates[[3]], dimensions = c(4,3), xlim = c(0, 8000), ylim = c(0, 8000), xlab = "", ylab = "")
A three dimensional plot of the same case is provided as a static image and can be obtained using the following code.
cytoPlot3d(templates.optimalFlow.barycenter$templates[[3]], dimensions = c(4, 3, 9), xlim = c(0, 8000), ylim = c(0, 8000), zlim = c(0, 8000))
cosa
We do the same for the density based hierarchical clustering.
cytoPlot(templates.optimalFlow.hdbscan$templates[[3]], dimensions = c(4,3), xlim = c(0, 8000), ylim = c(0, 8000), xlab = "", ylab = "")
cytoPlot3d(templates.optimalFlow.hdbscan$templates[[3]], dimensions = c(4, 3, 9), xlim = c(0, 8000), ylim = c(0, 8000), zlim = c(0, 8000))
From the visual representations of above we see that the different methods for obtaining a prototype cytometry for the cluster of cytometries labelled as 3 are returning similar results. Even more, results do seem to summarize in a reasonable way the information contained in the cytometries of cluster 3. Hence, doing some 2 or 3-dimensional visual inspection of the results is advisable for the user and it allows for an informed inspection of our procedures.
A totally unsupervised way of obtaining groups and templates is given by using density-based hierarchical clustering both when clustering the database of cytometries and when obtaining the prototype cytometry.
templates.optimalFlow.unsup <-
optimalFlowTemplates(
database = database, hclust.method = "hdbscan", cl.paral = 1, consensus.method = "hierarchical"
)## [1] "step 1: 18.5904183387756 secs"
## [1] "step 2: 1.46581335067749 mins"
## [1] "Execution time: 1.77565957307816 mins"print(templates.optimalFlow.unsup$clustering)## [1] 7 2 6 6 5 5 4 4 4 7 2 5 3 1 3print(templates.optimalFlow$clustering)## [1] 1 2 3 3 3 3 4 4 4 1 2 3 5 5 5cytoPlot(templates.optimalFlow.unsup$templates[[5]], dimensions = c(4,3), xlim = c(0, 8000), ylim = c(0, 8000), xlab = "", ylab = "")
4 optimalFlowClassification
Once we have a grouped database with prototype cytometries for each group we can apply different supervised classification procedures to classify a new ungated cytometry.
We start by selecting a test cytometry which we will treat as a cytometry that we want to classify in order to see how our supervised classification methods work.
test.cytometry <- Cytometry1[which(match(Cytometry1$`Population ID (name)`, c("Monocytes", "CD4+CD8-", "Mature SIg Kappa", "TCRgd-"), nomatch = 0) > 0), ]Let us begin with, essentially, the default method for using optimalFlowClassification. It consists in doing quadratic discriminant analysis using the most similar template.
classification.optimalFlow <-
optimalFlowClassification(
test.cytometry[, 1:10], database, templates.optimalFlow,
consensus.method = "pooling", cl.paral = 1
)## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.337722539901733 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.472489833831787 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.498994827270508 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.340389728546143 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.381137609481812 secs"
##
## [1] "step 1: 2.04118084907532 secs"## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## [1] "Similarity distances to templates:"
## [,1] [,2] [,3] [,4] [,5]
## [1,] 0.1480501 0.2994856 0.2598811 0.3117271 0.3368663
## [1] "step 2: 1.13986396789551 secs"
## [1] "step 3: 0.0531270503997803 secs"
## Time difference of 3.234862 secsIn order to execute properly our supervised classifications we need to provide the new cytometry we want to classify, in our case it is test.cytometry without the label information. Then we need to provide the database in which we have applied optimalFlowTemplates, the object returned by optimalFlowTemplates and the consensus.method that we have used in optimalFlowTemplates. This is necesary in order to be able to perform correctly the classification task.
The result we obtain is a list that we will analyse here to make the user familiar with it. The first argument is a clustering of the cytometry of interest.
head(classification.optimalFlow$cluster)## [1] CD4+CD8- CD4+CD8- CD4+CD8- CD4+CD8- CD4+CD8- CD4+CD8-
## Levels: CD4+CD8- Mature SIg Kappa Monocytes TCRgd-table(classification.optimalFlow$cluster)##
## CD4+CD8- Mature SIg Kappa Monocytes TCRgd-
## 7697 1577 6430 123The argument clusterings contains the initial unsupervized or semi-supervized clusterings of the cytometry of interest. It is itself a list that can have as much entries as the number of templates in the semi-supervized case, or only one entry in the case of initial.method = “unsupervized”. The relevant argument for the clusterings is cluster.
length(classification.optimalFlow$clusterings)## [1] 5table(classification.optimalFlow$clusterings[[1]]$cluster)##
## 1 2 3 4
## 7697 1577 6430 123As we see the initial clustering is not the same as the final result.
Finally, we have an entry that indicates which prototype is the closest to the new cytometry. This information is relevant since it is the prototype used for classifying in the default execution of optimalFlowClassification.
classification.optimalFlow$assigned.template.index## [1] 1templates.optimalFlow$clustering## [1] 1 2 3 3 3 3 4 4 4 1 2 3 5 5 5In this case test.cytometry is closest to the template corresponding to cluster 1 in templates.optimalFlow$clustering.
As we are performing supervised classification, a measure of how well our procedure works is in order. We have provided simple functions to calculate the median F-measure (see del Barrio et al. (2019) for details).
scoreF1.optimalFlow <- optimalFlow::f1Score(classification.optimalFlow$cluster,
test.cytometry, noise.types)
print(scoreF1.optimalFlow)## CD4+CD8- Mature SIg Kappa Monocytes TCRgd-
## F1-score 1 1 1 1
## Precision 1 1 1 1
## Recall 1 1 1 1We see that median F1-score, the first row in the table is close to 1 for each cell type, reflecting that classification is good for each cell type.
When using a consensus method that is not pooling, the default in optimalFlowTemplates, we have to assign cell types to the clusters in the prototype cytometries. This is done by voting, since our database is formed by gated cytometries and we have assigned cell types.
classification.optimalFlow.barycenter <-
optimalFlowClassification(
test.cytometry[, 1:10],
database, templates.optimalFlow.barycenter, consensus.method = "k-barycenter", cl.paral = 1
)## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.725910663604736 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.857559442520142 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.548777103424072 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.915315389633179 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.596104145050049 secs"
##
## [1] "step 1: 3.65867590904236 secs"## Weights do not add to 1. A normalization will be applied.## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## Weights do not add to 1. A normalization will be applied.## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## Weights do not add to 1. A normalization will be applied.## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## Weights do not add to 1. A normalization will be applied.## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## [1] "Similarity distances to templates:"
## [,1] [,2] [,3] [,4] [,5]
## [1,] 0.151584 0.2999437 0.2614572 0.3032183 0.3426324
## [1] "step 2: 0.176708221435547 secs"
## [1] "step 3: 0.0695259571075439 secs"
## Time difference of 3.905716 secstable(classification.optimalFlow.barycenter$cluster)##
## 1 2 3 4
## 123 1577 6430 7697classification.optimalFlow.barycenter$cluster.vote## $`1`
## cell simple.proportion
## 1 TCRgd- 1
##
## $`2`
## cell simple.proportion
## 1 Mature SIg Kappa 1
##
## $`3`
## cell simple.proportion
## 1 Monocytes 1
##
## $`4`
## cell simple.proportion
## 1 CD4+CD8- 1In this case the fuzzy classification is a hard one, since values of simple.proportion are all one. This means that label 1 is assigned to the entry in cell and so on.
classification.optimalFlow.barycenter$assigned.template.index## [1] 1templates.optimalFlow.barycenter$clustering## [1] 1 2 3 3 3 3 4 4 4 1 2 3 5 5 5Since usually the relabelling in classification.optimalFlow.barycenter$cluster.vote is fuzzy, we need to convert it to a hard clustering and then apply our median F-measure criteria. This is done as follows.
scoreF1.optimalFlow.barycenter <-
f1ScoreVoting(
classification.optimalFlow.barycenter$cluster.vote, classification.optimalFlow.barycenter$cluster,
test.cytometry,
1.01, noise.types
)
print(scoreF1.optimalFlow.barycenter$F1_score)## TCRgd- Mature SIg Kappa Monocytes CD4+CD8-
## F1-score 1 1 1 1
## Precision 1 1 1 1
## Recall 1 1 1 1Exactly the same applies for the case of templates.optimalFlow.hdbscan.
classification.optimalFlow.hdbscan <-
optimalFlowClassification(
test.cytometry[, 1:10],
database, templates.optimalFlow.hdbscan, consensus.method = "hierarchical", cl.paral = 1
)## [1] "tclust looking for k = 2"
## [1] "tclust found k = 1"
## [1] "0.769766092300415 secs"
##
## [1] "tclust looking for k = 2"
## [1] "tclust found k = 1"
## [1] "0.510528802871704 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "1.06597471237183 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.583372354507446 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.52520489692688 secs"
##
## [1] "step 1: 3.46953630447388 secs"## Weights do not add to 1. A normalization will be applied.## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## Weights do not add to 1. A normalization will be applied.## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## [1] "Similarity distances to templates:"
## [,1] [,2] [,3] [,4] [,5]
## [1,] 0.5278954 0.5701123 0.2598811 0.3117271 0.3368663
## [1] "step 2: 0.161943435668945 secs"
## [1] "step 3: 0.0475742816925049 secs"
## Time difference of 3.67983 secstable(classification.optimalFlow.hdbscan$cluster)##
## 1 2 3 4
## 1577 6430 124 7696classification.optimalFlow.hdbscan$cluster.vote## $`1`
## cell simple.proportion
## 1 Mature SIg Kappa 1
##
## $`2`
## cell simple.proportion
## 1 Monocytes 1
##
## $`3`
## cell simple.proportion
## 1 TCRgd- 1
##
## $`4`
## cell simple.proportion
## 1 CD4+CD8- 1classification.optimalFlow.hdbscan$assigned.template.index## [1] 3templates.optimalFlow.hdbscan$clustering## [1] 1 2 3 3 3 3 4 4 4 1 2 3 5 5 5scoreF1.optimalFlow.hdbscan <-
f1ScoreVoting(
classification.optimalFlow.hdbscan$cluster.vote, classification.optimalFlow.hdbscan$cluster,
test.cytometry,
1.01, noise.types
)
print(scoreF1.optimalFlow.hdbscan$F1_score)## Mature SIg Kappa Monocytes TCRgd- CD4+CD8-
## F1-score 1 1 0.9959514 0.9999350
## Precision 1 1 0.9919355 1.0000000
## Recall 1 1 1.0000000 0.9998701Another way of doing classification is to relabel the initial clustering that we obtain using the most similar template obtained by optimalFlowTemplates. This is called label-transfer and is further explained in del Barrio et al. (2019).
classification.optimalFlow.2 <-
optimalFlowClassification(
test.cytometry[, 1:10],
database, templates.optimalFlow, consensus.method = "pooling", classif.method = "matching",
cost.function = "ellipses", cl.paral = 1
)## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.537897348403931 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.952532052993774 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "1.10692095756531 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.715469360351562 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.553377389907837 secs"
##
## [1] "step 1: 3.90879273414612 secs"## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## [1] "Similarity distances to templates:"
## [,1] [,2] [,3] [,4] [,5]
## [1,] 0.1480501 0.2994856 0.2598811 0.3117271 0.3368663
## [1] "step 2: 0.290672302246094 secs"## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## [1] "0.0843331813812256 secs"
## [1] "step 3: 0.298254728317261 secs"
## Time difference of 4.498668 secstable(classification.optimalFlow.2$cluster)##
## 1 2 3 4
## 7697 1577 6430 123table(classification.optimalFlow.2$clusterings[[1]]$cluster)##
## 1 2 3 4
## 7697 1577 6430 123classification.optimalFlow.2$cluster.vote## $`1`
## cell compound.proportion simple.proportion
## 1 CD4+CD8- 1 1
##
## $`2`
## cell compound.proportion simple.proportion
## 1 Mature SIg Kappa 1 1
##
## $`3`
## cell compound.proportion simple.proportion
## 1 Monocytes 1 1
##
## $`4`
## cell compound.proportion simple.proportion
## 1 TCRgd- 1 1classification.optimalFlow.2$assigned.template.index## [1] 1templates.optimalFlow$clustering## [1] 1 2 3 3 3 3 4 4 4 1 2 3 5 5 5scoreF1.optimalFlow.2 <-
f1ScoreVoting(
classification.optimalFlow.2$cluster.vote, classification.optimalFlow.2$cluster,
test.cytometry,
1.01, noise.types
)
print(scoreF1.optimalFlow.2$F1_score)## CD4+CD8- Mature SIg Kappa Monocytes TCRgd-
## F1-score 1 1 1 1
## Precision 1 1 1 1
## Recall 1 1 1 1Also, classical techniques as random forest are available.
classification.optimalFlow.3 <-
optimalFlowClassification(
test.cytometry[, 1:10],
database, templates.optimalFlow, consensus.method = "pooling",
classif.method = "random forest", cl.paral = 1
)## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.696957349777222 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "1.09469151496887 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "1.09933066368103 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.714186668395996 secs"
##
## [1] "tclust looking for k = 4"
## [1] "tclust found k = 3"
## [1] "0.378019094467163 secs"
##
## [1] "step 1: 3.99740028381348 secs"## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## [1] "Similarity distances to templates:"
## [,1] [,2] [,3] [,4] [,5]
## [1,] 0.1480501 0.2994856 0.2598811 0.3117271 0.3368663
## [1] "step 2: 0.114541530609131 secs"## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'
## Warning in class(a) == class(b) || (is(a, "pgrid") && is(b, "wpp")): 'length(x)
## = 2 > 1' in coercion to 'logical(1)'## [1] "step 3: 26.63108086586 secs"
## Time difference of 30.74379 secstable(classification.optimalFlow.3$cluster)##
## CD4+CD8- Mature SIg Kappa Monocytes TCRgd-
## 7697 1577 6430 123classification.optimalFlow.3$assigned.template.index # the cytometry used for learning belongs to the cluster labelled as 1 and is the first of the cytometries in that cluster, hence it is the first cytometry in the database.## [1] 1 1templates.optimalFlow$clustering## [1] 1 2 3 3 3 3 4 4 4 1 2 3 5 5 5scoreF1.optimalFlow.3 <-
optimalFlow::f1Score(classification.optimalFlow.3$cluster,
test.cytometry,
noise.types
)
print(scoreF1.optimalFlow.3)## CD4+CD8- Mature SIg Kappa Monocytes TCRgd-
## F1-score 1 1 1 1
## Precision 1 1 1 1
## Recall 1 1 1 1References
del Barrio, Eustasio, Hristo Inouzhe, Jean-Michel Loubes, Agustin Mayo-Iscar, and Carlos Matran. 2019. “optimalFlow: Optimal-Transport Approach to Flow Cytometry Analysis,” July. https://arxiv.org/abs/1907.08006.
Garcia-Escudero, Luis-Angel, Alfonso Gordaliza, Carlos Matran, and Agustin Mayo-Iscar. 2008. “A General Trimming Approach to Robust Cluster Analysis.” The Annals of Statistics 36: 1324–45.