BootStrap runs for both scGPS training and prediction

ElasticNet and LDA prediction for each of all the subpopulations in the new mixed population after training the model for a subpopulation in the first mixed population. The number of bootstraps to be run can be specified.

bootstrap_prediction(
  nboots = 1,
  genes = genes,
  mixedpop1 = mixedpop1,
  mixedpop2 = mixedpop2,
  c_selectID = NULL,
  listData = list(),
  cluster_mixedpop1 = NULL,
  cluster_mixedpop2 = NULL,
  trainset_ratio = 0.5,
  LDA_run = TRUE,
  verbose = FALSE,
  log_transform = FALSE
)

Arguments

nboots	a number specifying how many bootstraps to be run
genes	a gene list to build the model
mixedpop1	a SingleCellExperiment object from a mixed population for training
mixedpop2	a SingleCellExperiment object from a target mixed population for prediction
c_selectID	the root cluster in mixedpop1 to becompared to clusters in mixedpop2
listData	a list object, which contains trained results for the first mixed population
cluster_mixedpop1	a vector of cluster assignment for mixedpop1
cluster_mixedpop2	a vector of cluster assignment for mixedpop2
trainset_ratio	a number specifying the proportion of cells to be part of the training subpopulation
LDA_run	logical, if the LDA prediction is added to compare to ElasticNet
verbose	a logical whether to display additional messages
log_transform	boolean whether log transform should be computed

Value

a list with prediction results written in to the index out_idx

See also

bootstrap_parallel for parallel options

Author

Quan Nguyen, 2017-11-25

Examples

day2 <- day_2_cardio_cell_sample
mixedpop1 <-new_scGPS_object(ExpressionMatrix = day2$dat2_counts, 
    GeneMetadata = day2$dat2geneInfo, CellMetadata = day2$dat2_clusters)
day5 <- day_5_cardio_cell_sample
mixedpop2 <-new_scGPS_object(ExpressionMatrix = day5$dat5_counts, 
    GeneMetadata = day5$dat5geneInfo, CellMetadata = day5$dat5_clusters)
genes <-training_gene_sample
genes <-genes$Merged_unique
cluster_mixedpop1 <- colData(mixedpop1)[,1]
cluster_mixedpop2 <- colData(mixedpop2)[,1]
c_selectID <- 2
test <- bootstrap_prediction(nboots = 1, mixedpop1 = mixedpop1, 
    mixedpop2 = mixedpop2, genes=genes, listData =list(), 
    cluster_mixedpop1 = cluster_mixedpop1, 
    cluster_mixedpop2 = cluster_mixedpop2, c_selectID = c_selectID)
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> Warning: variables are collinear
#> done training for bootstrap 1, moving to prediction...
names(test)
#> [1] "Accuracy"          "ElasticNetGenes"   "Deviance"         
#> [4] "ElasticNetFit"     "LDAFit"            "predictor_S1"     
#> [7] "ElasticNetPredict" "LDAPredict"        "cell_results"     
test$ElasticNetPredict
#> [[1]]
#> [[1]][[1]]
#> NULL
#> 
#> [[1]][[2]]
#> NULL
#> 
#> [[1]][[3]]
#> NULL
#> 
#> [[1]][[4]]
#> NULL
#> 
#> [[1]][[5]]
#> [1] "ElasticNet for subpop1 in target mixedpop2"
#> 
#> [[1]][[6]]
#> [1] 37.43316
#> 
#> [[1]][[7]]
#> [1] "ElasticNet for subpop2 in target mixedpop2"
#> 
#> [[1]][[8]]
#> [1] 32.14286
#> 
#> [[1]][[9]]
#> [1] "ElasticNet for subpop3 in target mixedpop2"
#> 
#> [[1]][[10]]
#> [1] 100
#> 
#> [[1]][[11]]
#> [1] "ElasticNet for subpop4 in target mixedpop2"
#> 
#> [[1]][[12]]
#> [1] 82.5
#> 
#> 
test$LDAPredict
#> [[1]]
#> [[1]][[1]]
#> NULL
#> 
#> [[1]][[2]]
#> NULL
#> 
#> [[1]][[3]]
#> NULL
#> 
#> [[1]][[4]]
#> NULL
#> 
#> [[1]][[5]]
#> [1] "LDA for subpop 1 in target mixedpop2"
#> 
#> [[1]][[6]]
#> [1] 94.65241
#> 
#> [[1]][[7]]
#> [1] "LDA for subpop 2 in target mixedpop2"
#> 
#> [[1]][[8]]
#> [1] 9.285714
#> 
#> [[1]][[9]]
#> [1] "LDA for subpop 3 in target mixedpop2"
#> 
#> [[1]][[10]]
#> [1] 87.96992
#> 
#> [[1]][[11]]
#> [1] "LDA for subpop 4 in target mixedpop2"
#> 
#> [[1]][[12]]
#> [1] 70
#> 
#>