Skip to content

Commit 9399fc2

Browse files
ajfedorecajfedorec
committed
plate reader calibration workflow upgrades
- seperated and updated parsing of calibration data into spark_parse() from generate_cfs() - removed hardcoded assumptions about fluorophores from calibration process
1 parent 2e84f81 commit 9399fc2

11 files changed

+344
-408
lines changed

‎DESCRIPTION

Lines changed: 4 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -1,7 +1,7 @@
11
Package: flopr
22
Type: Package
33
Title: A package for the normalisation and calibration of flow cytometry and plate reader spectrophotometry data
4-
Version: 0.1.0
4+
Version: 0.2.0
55
Authors@R: person("Fedorec", "Alex J H", , "[email protected]", c("aut", "cre"))
66
Author: Alex J H Fedorec [aut, cre]
77
Maintainer: Alex J H Fedorec <[email protected]>
@@ -25,5 +25,6 @@ Imports:
2525
rlang,
2626
tidyr,
2727
broom,
28-
mgcv
29-
RoxygenNote: 7.1.0
28+
mgcv,
29+
tidyselect
30+
RoxygenNote: 7.1.1

‎NAMESPACE

Lines changed: 1 addition & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1,13 +1,12 @@
11
# Generated by roxygen2: do not edit by hand
22

3-
export(generateCfs)
43
export(generate_cfs)
5-
export(prNorm)
64
export(process_fcs)
75
export(process_fcs_dir)
86
export(process_plate)
97
export(sparkParse)
108
export(spark_parse)
119
export(trim.fcs)
1210
importFrom(dplyr,"%>%")
11+
importFrom(rlang,":=")
1312
importFrom(rlang,.data)

‎R/miscFunctions.R

Lines changed: 103 additions & 53 deletions
Original file line numberDiff line numberDiff line change
@@ -2,80 +2,130 @@
22
#'
33
#' @param data_csv path to csv file from Tecan Spark plate reader
44
#' @param layout_csv path to csv file containing plate layout information
5+
#' @param timeseries Boolean flag indicating whether the data is a timeseries or
6+
#' single recording. The Tecan software outputs the two scenarios differently.
57
#'
68
#' @return a data.frame containing the parsed plate reader data
79
#' @export
810
#' @importFrom rlang .data
911
#'
10-
spark_parse <- function(data_csv, layout_csv) {
12+
spark_parse <- function(data_csv, layout_csv, timeseries=F) {
1113
data <- utils::read.table(data_csv, sep = ",", blank.lines.skip = T,
1214
header = F, stringsAsFactors = F)
1315

1416
plate_layout <- utils::read.csv(layout_csv)
1517

16-
start_time_idx <- which(data[, 1] == "Start Time")
17-
if (length(start_time_idx) > 1) {
18-
start_time_idx <- start_time_idx[length(start_time_idx)]
18+
if(timeseries == TRUE){
19+
start_time_idx <- which(data[, 1] == "Start Time")
20+
if (length(start_time_idx) > 1) {
21+
start_time_idx <- start_time_idx[length(start_time_idx)]
22+
}
23+
24+
next_block_start_idx <- start_time_idx + 2
25+
26+
end_of_file <- F
27+
all_data <- c()
28+
while (!end_of_file) {
29+
# find what is being measured
30+
block_name <- data[next_block_start_idx, 1]
31+
32+
# check if we are at the end of the file
33+
if (block_name == "End Time") {
34+
end_of_file <- T
35+
break
36+
}
37+
38+
# find where the end of the current measurement block is
39+
block_end_idx <- next_block_start_idx
40+
while (data[block_end_idx, 1] != "") {
41+
block_end_idx <- block_end_idx + 1
42+
}
43+
44+
# grab the data only for that measurement
45+
new_block <- data[(next_block_start_idx + 1):(block_end_idx - 1), ]
46+
47+
# trim unecessary readings i.e. temp and cycle number
48+
# and rename columns
49+
times <- new_block[2, ]
50+
new_block <- new_block[c(-1, -2, -3), ]
51+
names(new_block) <- times
52+
names(new_block)[1] <- "well"
53+
54+
# add info for each well
55+
joined_block <- dplyr::full_join(plate_layout, new_block)
56+
joined_block$measure <- block_name
57+
58+
#
59+
all_data <- rbind(all_data, joined_block)
60+
61+
#
62+
next_block_start_idx <- block_end_idx + 1
63+
}
64+
65+
# rearrange data ----------------------------------------------------------
66+
well_idx <- which(names(all_data) == "well")
67+
gathered_data <- tidyr::gather(all_data, key = "time", value = p,
68+
-c(1:well_idx, ncol(all_data)))
69+
gathered_data$time <- as.numeric(gathered_data$time)
70+
gathered_data$p <- as.numeric(gathered_data$p)
71+
spread_data <- tidyr::spread(gathered_data, key = .data$measure,
72+
value = .data$p)
73+
74+
spread_data$row <- substr(x = spread_data$well, start = 1, stop = 1)
75+
spread_data$column <- as.numeric(substr(x = spread_data$well, start = 2,
76+
stop = nchar(spread_data$well)))
77+
spread_data <- dplyr::arrange_at(spread_data, dplyr::vars(.data$time,
78+
.data$row,
79+
.data$column))
80+
81+
82+
83+
# write parsed data to csv ------------------------------------------------
84+
out_name <- gsub(".csv", "_parsed.csv", data_csv)
85+
utils::write.csv(x = spread_data, file = out_name, row.names = FALSE)
86+
87+
return(spread_data)
1988
}
89+
else if (timeseries == FALSE){
90+
start_time_idx <- which(data[, 1] == "Start Time") # get start and end time ids
91+
end_idx <- which(data[, 1] == "End Time")
92+
names_idx <- which(data[, 1] == "Name")
93+
names_idx <- names_idx[2:length(names_idx)] # remove the first start time entry which just details plate type
2094

21-
next_block_start_idx <- start_time_idx + 2
95+
all_data <- c()
96+
for (i in seq_len(length(start_time_idx))) {
97+
block_name <- data[names_idx[i], 2] # record name of what is being measured
2298

23-
end_of_file <- F
24-
all_data <- c()
25-
while (!end_of_file) {
26-
# find what is being measured
27-
block_name <- data[next_block_start_idx, 1]
99+
block_start <- start_time_idx[i] + 4 # find start and end of measurement block
100+
block_end_idx <- end_idx[i] - 3
28101

29-
# check if we are at the end of the file
30-
if (block_name == "End Time") {
31-
end_of_file <- T
32-
break
33-
}
102+
new_block <- data[(block_start):(block_end_idx), 1:2] # grab and name the data
103+
names(new_block)[1] <- "well"
104+
names(new_block)[2] <- "value"
34105

35-
# find where the end of the current measurement block is
36-
block_end_idx <- next_block_start_idx
37-
while (data[block_end_idx, 1] != "") {
38-
block_end_idx <- block_end_idx + 1
39-
}
106+
new_block$value <- as.numeric(new_block$value)
40107

41-
# grab the data only for that measurement
42-
new_block <- data[(next_block_start_idx + 1):(block_end_idx - 1), ]
108+
joined_block <- dplyr::full_join(plate_layout, new_block) # join to plate layout csv, add measurement category
109+
joined_block$measure <- block_name
43110

44-
# trim unecessary readings i.e. temp and cycle number
45-
# and rename columns
46-
times <- new_block[2, ]
47-
new_block <- new_block[c(-1, -2, -3), ]
48-
names(new_block) <- times
49-
names(new_block)[1] <- "well"
111+
all_data <- rbind(all_data, joined_block) # add to all data
112+
}
50113

51-
# add info for each well
52-
joined_block <- dplyr::full_join(plate_layout, new_block)
53-
joined_block$measure <- block_name
114+
# rearrange data ----------------------------------------------------------
115+
spread_data <- tidyr::pivot_wider(all_data, names_from = .data$measure,
116+
values_from = .data$value)
117+
spread_data$row <- substr(x = spread_data$well, start = 1, stop = 1)
118+
spread_data$column <- as.numeric(substr(x = spread_data$well, start = 2,
119+
stop = nchar(spread_data$well)))
120+
spread_data <- dplyr::arrange_at(spread_data, dplyr::vars(.data$row,
121+
.data$column))
54122

55-
#
56-
all_data <- rbind(all_data, joined_block)
123+
# write parsed data to csv ------------------------------------------------
124+
out_name <- gsub(".csv", "_parsed.csv", data_csv)
125+
utils::write.csv(x = spread_data, file = out_name, row.names = FALSE)
57126

58-
#
59-
next_block_start_idx <- block_end_idx + 1
127+
return(spread_data)
60128
}
61-
62-
# rearrange data ----------------------------------------------------------
63-
well_idx <- which(names(all_data) == "well")
64-
gathered_data <- tidyr::gather(all_data, key = "time", value = p,
65-
-c(1:well_idx, ncol(all_data)))
66-
gathered_data$time <- as.numeric(gathered_data$time)
67-
gathered_data$p <- as.numeric(gathered_data$p)
68-
spread_data <- tidyr::spread(gathered_data, key = .data$measure,
69-
value = .data$p)
70-
71-
spread_data$row <- substr(x = spread_data$well, start = 1, stop = 1)
72-
spread_data$column <- as.numeric(substr(x = spread_data$well, start = 2,
73-
stop = nchar(spread_data$well)))
74-
spread_data <- dplyr::arrange_at(spread_data, dplyr::vars(.data$time,
75-
.data$row,
76-
.data$column))
77-
78-
return(spread_data)
79129
}
80130

81131
#' Parser for Tecan Spark plate reader data

0 commit comments

Comments
 (0)