protGear vignette processing suite

Introduction

Analysis setup

Create 2 folders that hold the ‘.gpr’ files and the corresponding sample identifier files.

Folder structure of array and sample ID files

Sample identifier file

sample ID file structure

## install from BioConductor
if (!require("BiocManager", quietly = TRUE)) {
  install.packages("BiocManager")
}
BiocManager::install("protGear")

Importing data

## load the package
suppressWarnings(library(protGear))

The first part is to specify the parameters of the micro-array experiment to assist in processing the data. The parameters specified are

• channel - The scanner fluorescence output used to record the data. It can be green,red,blue among others with a specified number.
• chip_path - the folder where the sub folders of ‘gpr’ or ‘txt’ files are stored. This path contains sub folders with the array data, for example the sub folders for the different batches.
• totsamples - the number of samples in a slide or array.
  
• blockspersample - The number of blocks a sample it takes. In this example each sample occupies 2 blocks each with 384 spots.
• sampleID_path - the folder where the sample identifiers files are stored
• machine - The indicator for which machine was used to hybridize the samples if the experiment had more than one machine.
• date_process -the date of sample processing

The parameters “chip_path”, “channel” , “totsamples” and “sampleID_path” are mandatory.

## specify the the parameters to process the data
genepix_vars <- array_vars(
  channel = "635",
  chip_path = system.file("extdata/array_data/", package = "protGear"),
  totsamples = 21,
  blockspersample = 2,
  sampleID_path = system.file("extdata/array_sampleID/", package = "protGear"),
  mig_prefix = "_first",
  machine = 1,
  ## optional
  date_process = "0520"
)

The exact channel used should be checked in the header of the file from the quantification software under Wavelengths.

header_gpr <- readLines(system.file("extdata/array_data/machine1/KK2-06.txt", package = "protGear"),
  n = 40
)
header_gpr <- gsub("\"", "", header_gpr[1:32])
header_gpr[1:32]

##  [1] "ATF\t1\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                                
##  [2] "29\t31\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                                
##  [3] "Type=GenePix Export 3\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                               
##  [4] "DateTime=2017/09/05 17:17:47\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                        
##  [5] "Settings=\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                           
##  [6] "GalFile=C:\\Users\\rkimathi\\Desktop\\GAL FILE DATA.gal\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                             
##  [7] "PixelSize=10\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                        
##  [8] "Wavelengths=635\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                     
##  [9] "ImageFiles=Z:\\Multicenter study\\4_9_2017\\data\\BRB0001.tif 0\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                     
## [10] "NormalizationMethod=None\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                            
## [11] "NormalizationFactors=1\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                              
## [12] "JpegImage=\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                          
## [13] "StdDev=Type 1\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                       
## [14] "FeatureType=Circular\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                
## [15] "Barcode=\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                            
## [16] "BackgroundSubtraction=LocalFeature\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                  
## [17] "ImageOrigin=0, 0\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                    
## [18] "JpegOrigin=1670, 840\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                
## [19] "Creator=GenePix Pro 7.3.0.0\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                         
## [20] "Scanner=GenePix 4300 [141110]\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                       
## [21] "FocusPosition=0\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                     
## [22] "Temperature=0\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                       
## [23] "LinesAveraged=1\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                     
## [24] "Comment=\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                            
## [25] "PMTGain=400\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                         
## [26] "ScanPower=100\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                       
## [27] "LaserPower=0.55\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                     
## [28] "Filters=Standard Red\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                
## [29] "ScanRegion=0,0,2200,7200\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                            
## [30] "ScanArea=1\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                          
## [31] "Supplier=\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"                                                                                           
## [32] "Flags\tNormalize\tAutoflag\tBlock\tColumn\tRow\tName\tID\tX\tY\tF635 Median\tB635\tB635 Median\tSNR 635\tF635 Median - B635\tIndex"

The function check_sampleID_files() helps to check whether each micro array file has a corresponding sample identifier file. The sample identifier files are generated from the lab plate maps to match the corresponding samples on a specific slide.If the sample identifier file is missing, protGear automatically generates the id’s.

visualize_slide(
  infile = system.file("extdata/array_data/machine1/KK2-06.txt", package = "protGear"),
  MFI_var = "F635 Median"
)

visualize_slide_2d(
  infile = system.file("extdata/array_data/machine1/KK2-06.txt", package = "protGear"),
  MFI_var = "B635 Median"
)

#### read in all the datasets
### list all the file names under data folder
filenames <- list.files(file.path(genepix_vars$paths[[1]]),
  pattern = "*.txt$|*.gpr$", full.names = FALSE
)
#' @___________________read_in_the_files_with_the_text_data_from_the_chip_____________________________
### read all the data files and save them in a list
data_path <- paste0(genepix_vars$paths[[1]], "/")
data_files <- purrr::map(
  .x = filenames,
  .f = read_array_files,
  data_path = data_path,
  genepix_vars = genepix_vars
)
data_files <- set_names(data_files, purrr::map(filenames, name_of_files))

Background Correction

Background noise is caused by non-specific fluorescence such as auto-fluorescence of the glass slide or non-specific binding of parameters and reference substances. To cut down the effect of background noise we have included different approaches for background correction. First, we extract the background values, visualize and select the best background approach. We have implemented five different approaches;

1. Local background subtraction
2. Global background subtraction
3. Moving minimum background subtraction
4. Normal and exponential model (normexp)
5. Log-linear background correction (Edwards)

In ‘.gpr’ files the Background column starts with a ‘B’ followed by the wavelength or channel. In order to perform background correction, we extract the background mean fluorescent intensities (MFI’s) using the function extract_bg() . The function accepts the arguments iden which is the file identifier, data_files a list of data objects with names utilised by iden and genepix_vars defined using array_vars() function. We utilise the purr::map function to extract the background MFI of multiple data files.

## utilising the map package we process a number of files  under data_files list
dfs <- names(data_files)
allData_bg <- purrr::map(.x = dfs, .f = extract_bg, data_files = data_files, genepix_vars)
allData_bg <- set_names(allData_bg, purrr::map(filenames, name_of_files))
allData_bg <- plyr::ldply(allData_bg)

p1 <- plot_FB(allData_bg,
  antigen_name = "antigen",
  bg_MFI = "BG_Median", FG_MFI = "FBG_Median", log = FALSE
)


p1

p2 <- plot_bg(
  df = allData_bg, x_axis = "Block", bg_MFI = "BG_Median",
  log_mfi = TRUE
)
p2

sample_ID_merged_dfs <- purrr::map(
  .x = dfs, .f = merge_sampleID, data_files = data_files,
  genepix_vars, method = "subtract_local"
)
sample_ID_merged_dfs <- set_names(sample_ID_merged_dfs, purrr::map(filenames, name_of_files))

Buffer spots

Buffer spots are specific to the experiment design and are not always included. Buffer spots are used to check for unexpected scanning artefacts. The buffer spots should have similar values in different slides. Some outliers are expected for buffer spots close sample spots or landmark. However you can specify the name of your control antigens here buffer_spot=“buffer” if you do not have buffer spots.

buffer_transp <- purrr::map(.x = sample_ID_merged_dfs, .f = buffer_spots, buffer_spot = "buffer")

buffer_transp <- set_names(buffer_transp, purrr::map(filenames, name_of_files))

buffers <- plyr::ldply(buffer_transp)
plot_buffer(buffers, buffer_names = "antigen", buffer_mfi = "FMedianBG_correct", slide_id = ".id")

Coefficient of Variation (CV)

To calculate the CV’s we utilise the cv_estimation() function with a cv_cut_off specified , sample identifier variable and antigen specified under sampleID_var and antigen respectively. The replicate_var and mfi_var identifies the variable with the replicate rank generated and MFI’s values.

#' @________________________________calculated_cv_for_each_data_file_______________________________________
#' data without the selected mean for the best 2 CVs
dataCV <- purrr::map(
  .x = sample_ID_merged_dfs, .f = cv_estimation, lab_replicates = 3,
  cv_cut_off = 20,
  sampleID_var = "sampleID", antigen_var = "antigen", replicate_var = "replicate",
  mfi_var = "FMedianBG_correct"
)

lab_replicates <- 3
dataCV <- set_names(dataCV, purrr::map(filenames, name_of_files))

aa <- plyr::ldply(dataCV)
GGally::ggpairs(aa, aes(color = cvCat_all),
  columns = paste(seq_len(lab_replicates)), title = "", axisLabels = "show"
) +
  theme_light()

#' @________________________________summary_of_cv_for_each_sample________________________________________
#' creates summaries by cv's greater than 20 and less than 20

cv_cut_off <- 20
dataCV_sample <- purrr::map(
  .x = dataCV, .f = protGear::cv_by_sample_estimation, cv_variable = "cvCat_all",
  lab_replicates = 3
)
dataCV_sample <- set_names(dataCV_sample, purrr::map(filenames, name_of_files))
all_cv_sample <- plyr::ldply(dataCV_sample)

less_20 <- rlang::sym(paste0("CV <= ", cv_cut_off, "_perc"))
gt_20 <- rlang::sym(paste0("CV > ", cv_cut_off, "_perc"))

less_20_per <- rlang::sym(paste0("% CV <=", cv_cut_off))
gt_20_per <- rlang::sym(paste0("% CV >", cv_cut_off))
ggplot(all_cv_sample) +
  geom_violin(aes(.id, `CV <= 20_perc`, color = "% CV =<20")) +
  geom_violin(aes(.id, `CV > 20_perc`, color = "% CV >20")) +
  geom_violin(aes(.id, Others_perc, color = "Others")) +
  ylab("% of CV") +
  theme_minimal() +
  ggtitle("% of CV >20 or <=20 for each slide all repeats considered")

#' @________________________________data_with_selected_best_2_CV_______________________________________
#' data with the selected mean for the best 2 CVs
dataCV_best2 <- purrr::map(
  .x = dataCV, .f = best_CV_estimation, slide_id = "iden", lab_replicates = 3,
  cv_cut_off = 20
)

## give the names to the returned list
dataCV_best2 <- set_names(dataCV_best2, purrr::map(filenames, name_of_files))


dataCV_sample_best2 <- purrr::map(
  .x = dataCV_best2, .f = cv_by_sample_estimation,
  cv_variable = "best_CV_cat", lab_replicates = 3
)
dataCV_sample_best2 <- set_names(dataCV_sample_best2, purrr::map(filenames, name_of_files))
all_cv_sample_best2 <- plyr::ldply(dataCV_sample_best2)

## plot only the CV perccentages
ggplot(all_cv_sample_best2) +
  geom_violin(aes(.id, `CV <= 20_perc`, color = "% CV =<20")) +
  geom_violin(aes(.id, `CV > 20_perc`, color = "% CV >20")) +
  geom_violin(aes(.id, Others_perc, color = "Others")) +
  ylab("% of CV") +
  theme_minimal() +
  ggtitle("% of CV >20 or <=20 for each slide")

Tag subtraction

Tag subtraction is applied for antigens containing purification tags. A file with the specific TAG name for each antigen is loaded. The file has the antigen,TAG and TAG_name and the TAG_name must be part of the antigens listed.

tag_file <- read.csv(system.file("extdata/TAG_antigens.csv", package = "protGear"))
tag_antigens <- c("CD4TAG", "GST", "MBP")
batch_vars <- list(machine = "m1", day = "0520")

tb1 <- data.frame(head(tag_file, n = 10))
tb1 %>%
  knitr::kable()

#' @________________________________subtract_the_tag_values_______________________________________
#'
## tag subtract
## read in the KILCHip TAG file to substract GST-1, MBP -2 and  CD4TAG - 0 file
dataCV_tag <- purrr::map(
  .x = dataCV_best2, .f = tag_subtract,
  tag_antigens = tag_antigens, mean_best_CV_var = "mean_best_CV", tag_file = tag_file,
  antigen_var = "antigen",
  batch_vars = batch_vars
)
dataCV_tag <- set_names(dataCV_tag, purrr::map(filenames, name_of_files))
dataCV_tag <- plyr::ldply(dataCV_tag)

aaa <- dataCV_tag %>%
  filter(TAG_name == "GST")

aaa <- aaa %>%
  dplyr::select(.id, sampleID, antigen, mean_best_CV, mean_best_CV_tag)

aaa <- aaa %>%
  gather(measure, mfi, -c(.id:antigen))

ggplot(aaa, aes(as.factor(antigen), mfi, color = measure)) +
  geom_boxplot(aes(fill = measure), alpha = 0.5) +
  theme_light() +
  xlab("antigen name") +
  ggtitle("Before and after TAG subtraction") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

Normalisation

To normalise the data, we create a matrix matrix_antigen with all the data combined. We ensure the slide identifier is included as row names of the matrix or we have a way to join them after normalisation. The array_matrix matrix is used to hold these parameters. In place of AMA1 in the example below you use one of your features or antigen. We have implemented four different normalisation techniques;

1. Log₂ normalisation
2. Loess normalisation
3. RLM normalisation
4. VSN normalisation.

df_to_normalise <- dataCV_tag %>%
  ungroup() %>%
  dplyr::select(slide = .id, sampleID, sample_array_ID, antigen, mean_best_CV) %>%
  group_by(sampleID, slide)
df_to_normalise$sample_index <- group_indices(.data = df_to_normalise)

###
to_normalise <- df_to_normalise %>%
  ungroup() %>%
  dplyr::select(-slide, -sampleID, -sample_array_ID) %>%
  dplyr::select(antigen, sample_index, everything()) %>%
  gather(variable, value, -(antigen:sample_index)) %>%
  unite(temp, antigen) %>%
  dplyr::select(-variable) %>%
  spread(temp, value) %>%
  as.data.frame(.)

### get the row names of the machine data
row.names(to_normalise) <- to_normalise$sample_index
# batch_all <- as.factor(paste0(to_normalise$machine,"/",to_normalise$day))
# machines <- as.factor(to_normalise$machine)
# day_batches <- as.factor(to_normalise$day)

## create the matrix to normalise
matrix_antigen <- to_normalise %>%
  dplyr::select(-sample_index) %>%
  as.matrix(.)


## create the matrix to hold the important parameters
## in place of AMA1 you use one of your features or antigen
array_matrix <- df_to_normalise %>%
  filter(antigen == "AMA1") %>%
  ungroup() %>%
  dplyr::select(sample_array_ID, sample_index, slide)

control_antigens <- c("CommercialHumanIgG", "CD4TAG")

The matrix_normalise() function is used to normalise the data and returns a list of plots and a matrix of normalised values. A plot is returned if return_plot = TRUE otherwise only a matrix of normalised values will be returned.

normlise_df <- matrix_normalise(matrix_antigen,
  method = "vsn", array_matrix = array_matrix,
  return_plot = TRUE, control_antigens = control_antigens
)

normlise_df$plot_normalisation

control_antigens <- c("CommercialHumanIgG", "CD4TAG")
## no normalisation
normalise_list_none <- matrix_normalise(
  matrix_antigen = matrix_antigen,
  method = "none",
  array_matrix = array_matrix,
  return_plot = TRUE,
  control_antigens = control_antigens
)
names(normalise_list_none) <- c("matrix_antigen_none", "plot_none")
## log2 normalisation
normalise_list_log <- matrix_normalise(
  matrix_antigen = matrix_antigen,
  method = "log2",
  array_matrix = array_matrix,
  return_plot = TRUE,
  control_antigens = control_antigens
)
names(normalise_list_log) <- c("matrix_antigen_log", "plot_log")
## vsn normalisation
normalise_list_vsn <- matrix_normalise(
  matrix_antigen = matrix_antigen,
  method = "vsn",
  array_matrix = array_matrix,
  return_plot = TRUE,
  control_antigens = control_antigens
)
names(normalise_list_vsn) <- c("matrix_antigen_vsn", "plot_vsn")
## cyclic loess with log
normalise_list_cyclic_loess_log <- matrix_normalise(
  matrix_antigen = matrix_antigen,
  method = "cyclic_loess_log",
  array_matrix = array_matrix,
  return_plot = TRUE,
  control_antigens = control_antigens
)
names(normalise_list_cyclic_loess_log) <- c(
  "matrix_antigen_cyclic_loess_log",
  "plot_cyclic_loess_log"
)


normalise_list_rlm <- matrix_normalise(
  matrix_antigen = matrix_antigen,
  method = "rlm",
  array_matrix = array_matrix,
  return_plot = TRUE,
  control_antigens = control_antigens
)
names(normalise_list_rlm) <- c("matrix_antigen_rlm", "plot_rlm")


## create a list after normalisation
normalised_list <- c(
  normalise_list_none,
  normalise_list_log,
  normalise_list_vsn,
  normalise_list_cyclic_loess_log,
  normalise_list_rlm
)
##
normalised_list_plot <- normalised_list[grepl("plot", names(normalised_list))]

p <- do.call("grid.arrange", c(normalised_list_plot, ncol = 2))

Heatmaps

norm_df <- normlise_df$matrix_antigen_normalised
norm_df <- norm_df %>%
  dplyr::select(-control_antigens)
p3 <- pheatmap::pheatmap(norm_df,
  scale = "none", cluster_rows = FALSE,
  main = paste("VSN", "Normalised Data"),
  silent = TRUE
)
#-------
## if you want to save the file
# p3 <- ggplotify::as.ggplot(p3)
# p <- p3 +  theme_void()
# ggsave(p ,
#          filename ="heatmap.PNG" ,
#          width = 16 , height = 12 ,
#          limitsize = FALSE,
#          dpi=200 )
#-------
p3

PCA analysis

norm_df <- normlise_df$matrix_antigen_normalised
res_pca <- prcomp(norm_df, scale = TRUE)
var <- get_pca_var(res_pca)
vars_visualise <- 20
# Visualize the PCA
## individuals contributing to the PCA
p1 <- fviz_pca_ind(res_pca,
  col.var = "contrib", # Color by contributions to the PC
  gradient.cols = c("#00AFBB", "#E7B800", "#FC4E07"),
  repel = TRUE # Avoid text overlapping
) + theme_minimal()

# Select the top vars_visualise contributing variables
p2 <- fviz_pca_biplot(res_pca,
  label = "var",
  select.var = list(contrib = vars_visualise)
) +
  theme_minimal()


# Total cos2 of variables on Dim.1 and Dim.2
p3 <- fviz_cos2(res_pca, choice = "var", axes = 1:2, top = vars_visualise)


# Color by cos2 values: quality on the factor map
p4 <- fviz_pca_var(res_pca,
  col.var = "cos2",
  gradient.cols = c("#00AFBB", "#E7B800", "#FC4E07"),
  select.var = list(contrib = vars_visualise),
  repel = TRUE # Avoid text overlapping
)


## combine the plots into one grid
## combine the plots into one grid
## combine the plots into one grid
p_pca <- gridExtra::grid.arrange(p1, p2, p3, p4, ncol = 2)

## If you want to save the file
# ggsave(p_pca ,
#          filename ="p_pca.PNG" ,
#          width = 16 , height = 12 ,
#           units = "in",
#          limitsize = FALSE,
#          dpi=300)
p_pca

## TableGrob (2 x 2) "arrange": 4 grobs
##   z     cells    name           grob
## 1 1 (1-1,1-1) arrange gtable[layout]
## 2 2 (1-1,2-2) arrange gtable[layout]
## 3 3 (2-2,1-1) arrange gtable[layout]
## 4 4 (2-2,2-2) arrange gtable[layout]

Shiny application

Shiny is an R package that makes it easy to build interactive web apps straight from R. protGear has a built in user-friendly Shiny dashboard to assist in real-time processing and visualization. It provides five sequential steps for handling a data table of fluorescent intensities. The dashboard is launched as below from R or RStudio. To Install R: Open an internet browser and go to www.r-project.org.

protGear::launch_protGear_interactive()

sessionInfo()

## R version 4.4.2 (2024-10-31)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.1 LTS
## 
## Matrix products: default
## BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 
## LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so;  LAPACK version 3.12.0
## 
## locale:
##  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
##  [3] LC_TIME=en_US.UTF-8        LC_COLLATE=C              
##  [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
##  [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
##  [9] LC_ADDRESS=C               LC_TELEPHONE=C            
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       
## 
## time zone: Etc/UTC
## tzcode source: system (glibc)
## 
## attached base packages:
## [1] grid      stats     graphics  grDevices utils     datasets  methods  
## [8] base     
## 
## other attached packages:
##  [1] protGear_1.11.0     vsn_3.75.0          Biobase_2.67.0     
##  [4] BiocGenerics_0.53.3 generics_0.1.3      limma_3.63.2       
##  [7] tidyr_1.3.1         dplyr_1.1.4         remotes_2.5.0      
## [10] ggplotify_0.1.2     GGally_2.2.1        rlang_1.1.4        
## [13] magick_2.8.5        factoextra_1.0.7    FactoMineR_2.11    
## [16] styler_1.10.3       knitr_1.49          png_0.1-8          
## [19] gridExtra_2.3       data.table_1.16.2   pheatmap_1.0.12    
## [22] scales_1.3.0        purrr_1.0.2         gtools_3.9.5       
## [25] ggpubr_0.6.0        ggplot2_3.5.1      
## 
## loaded via a namespace (and not attached):
##   [1] RColorBrewer_1.1-3      sys_3.4.3               jsonlite_1.8.9         
##   [4] magrittr_2.0.3          estimability_1.5.1      farver_2.1.2           
##   [7] rmarkdown_2.29          fs_1.6.5                zlibbioc_1.52.0        
##  [10] vctrs_0.6.5             memoise_2.0.1           rstatix_0.7.2          
##  [13] htmltools_0.5.8.1       broom_1.0.7             Formula_1.2-5          
##  [16] gridGraphics_0.5-1      sass_0.4.9              bslib_0.8.0            
##  [19] htmlwidgets_1.6.4       plyr_1.8.9              plotly_4.10.4          
##  [22] emmeans_1.10.5          cachem_1.1.0            buildtools_1.0.0       
##  [25] mime_0.12               lifecycle_1.0.4         pkgconfig_2.0.3        
##  [28] Matrix_1.7-1            R6_2.5.1                fastmap_1.2.0          
##  [31] GenomeInfoDbData_1.2.13 MatrixGenerics_1.19.0   shiny_1.9.1            
##  [34] digest_0.6.37           colorspace_2.1-1        AnnotationDbi_1.69.0   
##  [37] S4Vectors_0.45.2        RSQLite_2.3.8           labeling_0.4.3         
##  [40] fansi_1.0.6             mgcv_1.9-1              httr_1.4.7             
##  [43] abind_1.4-8             compiler_4.4.2          bit64_4.5.2            
##  [46] withr_3.0.2             backports_1.5.0         carData_3.0-5          
##  [49] DBI_1.2.3               ggstats_0.7.0           R.utils_2.12.3         
##  [52] ggsignif_0.6.4          MASS_7.3-61             scatterplot3d_0.3-44   
##  [55] flashClust_1.01-2       tools_4.4.2             httpuv_1.6.15          
##  [58] R.oo_1.27.0             glue_1.8.0              nlme_3.1-166           
##  [61] R.cache_0.16.0          promises_1.3.2          cluster_2.1.6          
##  [64] gtable_0.3.6            tzdb_0.4.0              R.methodsS3_1.8.2      
##  [67] preprocessCore_1.69.0   hms_1.1.3               car_3.1-3              
##  [70] utf8_1.2.4              XVector_0.47.0          ggrepel_0.9.6          
##  [73] pillar_1.9.0            yulab.utils_0.1.8       genefilter_1.89.0      
##  [76] later_1.4.1             splines_4.4.2           Kendall_2.2.1          
##  [79] lattice_0.22-6          survival_3.7-0          bit_4.5.0              
##  [82] annotate_1.85.0         tidyselect_1.2.1        maketools_1.3.1        
##  [85] Biostrings_2.75.1       IRanges_2.41.1          shinydashboard_0.7.2   
##  [88] stats4_4.4.2            xfun_0.49               statmod_1.5.0          
##  [91] matrixStats_1.4.1       DT_0.33                 UCSC.utils_1.3.0       
##  [94] lazyeval_0.2.2          yaml_2.3.10             boot_1.3-31            
##  [97] evaluate_1.0.1          tibble_3.2.1            BiocManager_1.30.25    
## [100] multcompView_0.1-10     cli_3.6.3               affyio_1.77.0          
## [103] xtable_1.8-4            munsell_0.5.1           jquerylib_0.1.4        
## [106] Rcpp_1.0.13-1           GenomeInfoDb_1.43.2     XML_3.99-0.17          
## [109] leaps_3.2               readr_2.1.5             blob_1.2.4             
## [112] flexdashboard_0.6.2     viridisLite_0.4.2       mvtnorm_1.3-2          
## [115] affy_1.85.0             crayon_1.5.3            KEGGREST_1.47.0