Cowpea Screen 04
This notebook describes the data analysis pipeline for side-view images collected for 4th cowpea screen (Cowpea 04), performed between April 15th and May 19th, 2022 by Hayley Sussman. The data was analyzed using the PlantCV pipeline by Magda Julkowska, and used here as an input.
Collating all side-views into one value for each plant
Since all of the plants were imaged from 7 angles, we first need to summarize the total plant size, by summarizing the pixels assigned as “plant” from all different side views. We will do it based on the time-stamp, for all of the imaging campaigns.
Let’s establish the pipeline first for the first file:
my_files <- list.files(pattern = "traits.csv")
length(my_files)## [1] 8data <- read.csv(my_files[1])
my_data <- data[,c(31, 18:25,27:30)]
my_dataLet’s split the ROI into specific information - including the timestamp!
for(i in 1:nrow(my_data)){
my_data$timestamp[i] <- strsplit(my_data$roi[i], "_")[[1]][1]
}
my_data#install.packages("doBy")
library(doBy)
sum_data <- summaryBy(area + convex_hull_area + solidity + perimeter + width + height ~ timestamp, data = my_data, FUN = function(x) c(mean = mean(x), sum=sum(x), max=max(x)))
sum_datalet’s keep only the information we find valuable:
colnames(sum_data)## [1] "timestamp" "area.mean" "area.sum"
## [4] "area.max" "convex_hull_area.mean" "convex_hull_area.sum"
## [7] "convex_hull_area.max" "solidity.mean" "solidity.sum"
## [10] "solidity.max" "perimeter.mean" "perimeter.sum"
## [13] "perimeter.max" "width.mean" "width.sum"
## [16] "width.max" "height.mean" "height.sum"
## [19] "height.max"sum_clean <- sum_data[,c(1, 3, 5, 8, 11, 16, 19)]
sum_cleanGreat - before moving forward - let’s add also info on DAY:
for(i in 1:nrow(sum_clean)){
sum_clean$day[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][1]
sum_clean$time[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanFor this experiment - I see in the lab notebook that “Pots missed for imaging on May 19th: 47 and 127”
So - we should have complete set of pots
So now we need to make a list of numbers from 1-140 and remove these missing numbers and merge them inside this data:
pots <- 1:140
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
## [19] 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
## [37] 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
## [55] 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
## [73] 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
## [91] 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108
## [109] 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126
## [127] 127 128 129 130 131 132 133 134 135 136 137 138 139 140sum_clean$pot.no <- pots
sum_cleanCP04 <- sum_cleanOK - now let’s do next day but much quicker:
data <- read.csv(my_files[2])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$timestamp[i] <- strsplit(my_data$roi[i], "_")[[1]][1]
}
sum_data <- summaryBy(area + convex_hull_area + solidity + perimeter + width + height ~ timestamp, data = my_data, FUN = function(x) c(mean = mean(x), sum=sum(x), max=max(x)))
sum_clean <- sum_data[,c(1, 3, 5, 8, 11, 16, 19)]
for(i in 1:nrow(sum_clean)){
sum_clean$day[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][1]
sum_clean$time[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanpots <- 1:140
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
## [19] 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
## [37] 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
## [55] 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
## [73] 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
## [91] 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108
## [109] 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126
## [127] 127 128 129 130 131 132 133 134 135 136 137 138 139 140sum_clean$pot.no <- pots
sum_cleanCP04 <- rbind(CP04, sum_clean)data <- read.csv(my_files[3])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$timestamp[i] <- strsplit(my_data$roi[i], "_")[[1]][1]
}
sum_data <- summaryBy(area + convex_hull_area + solidity + perimeter + width + height ~ timestamp, data = my_data, FUN = function(x) c(mean = mean(x), sum=sum(x), max=max(x)))
sum_clean <- sum_data[,c(1, 3, 5, 8, 11, 16, 19)]
for(i in 1:nrow(sum_clean)){
sum_clean$day[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][1]
sum_clean$time[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanpots <- 1:140
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
## [19] 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
## [37] 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
## [55] 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
## [73] 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
## [91] 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108
## [109] 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126
## [127] 127 128 129 130 131 132 133 134 135 136 137 138 139 140sum_clean$pot.no <- pots
sum_cleanCP04 <- rbind(CP04, sum_clean)data <- read.csv(my_files[4])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$timestamp[i] <- strsplit(my_data$roi[i], "_")[[1]][1]
}
sum_data <- summaryBy(area + convex_hull_area + solidity + perimeter + width + height ~ timestamp, data = my_data, FUN = function(x) c(mean = mean(x), sum=sum(x), max=max(x)))
sum_clean <- sum_data[,c(1, 3, 5, 8, 11, 16, 19)]
for(i in 1:nrow(sum_clean)){
sum_clean$day[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][1]
sum_clean$time[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanpots <- 1:140
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
## [19] 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
## [37] 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
## [55] 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
## [73] 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
## [91] 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108
## [109] 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126
## [127] 127 128 129 130 131 132 133 134 135 136 137 138 139 140sum_clean$pot.no <- pots
sum_cleanCP04 <- rbind(CP04, sum_clean)data <- read.csv(my_files[5])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$timestamp[i] <- strsplit(my_data$roi[i], "_")[[1]][1]
}
sum_data <- summaryBy(area + convex_hull_area + solidity + perimeter + width + height ~ timestamp, data = my_data, FUN = function(x) c(mean = mean(x), sum=sum(x), max=max(x)))
sum_clean <- sum_data[,c(1, 3, 5, 8, 11, 16, 19)]
for(i in 1:nrow(sum_clean)){
sum_clean$day[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][1]
sum_clean$time[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanpots <- 1:140
sum_clean$pot.no <- pots
sum_cleanCP04 <- rbind(CP04, sum_clean)data <- read.csv(my_files[6])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$timestamp[i] <- strsplit(my_data$roi[i], "_")[[1]][1]
}
sum_data <- summaryBy(area + convex_hull_area + solidity + perimeter + width + height ~ timestamp, data = my_data, FUN = function(x) c(mean = mean(x), sum=sum(x), max=max(x)))
sum_clean <- sum_data[,c(1, 3, 5, 8, 11, 16, 19)]
for(i in 1:nrow(sum_clean)){
sum_clean$day[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][1]
sum_clean$time[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanIn here - I figured that pots 49 and 50 were missing - by looking at the images.
pots <- 1:140
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
## [19] 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
## [37] 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
## [55] 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
## [73] 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
## [91] 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108
## [109] 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126
## [127] 127 128 129 130 131 132 133 134 135 136 137 138 139 140sum_clean$pot.no <- pots
sum_cleanCP04 <- rbind(CP04, sum_clean)data <- read.csv(my_files[7])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$timestamp[i] <- strsplit(my_data$roi[i], "_")[[1]][1]
}
sum_data <- summaryBy(area + convex_hull_area + solidity + perimeter + width + height ~ timestamp, data = my_data, FUN = function(x) c(mean = mean(x), sum=sum(x), max=max(x)))
sum_clean <- sum_data[,c(1, 3, 5, 8, 11, 16, 19)]
for(i in 1:nrow(sum_clean)){
sum_clean$day[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][1]
sum_clean$time[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanpots <- 1:140
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
## [19] 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
## [37] 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
## [55] 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
## [73] 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
## [91] 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108
## [109] 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126
## [127] 127 128 129 130 131 132 133 134 135 136 137 138 139 140sum_clean$pot.no <- pots
sum_cleanCP04 <- rbind(CP04, sum_clean)data <- read.csv(my_files[8])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$timestamp[i] <- strsplit(my_data$roi[i], "_")[[1]][1]
}
sum_data <- summaryBy(area + convex_hull_area + solidity + perimeter + width + height ~ timestamp, data = my_data, FUN = function(x) c(mean = mean(x), sum=sum(x), max=max(x)))
sum_clean <- sum_data[,c(1, 3, 5, 8, 11, 16, 19)]
for(i in 1:nrow(sum_clean)){
sum_clean$day[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][1]
sum_clean$time[i] <- strsplit(sum_clean$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanpots <- 1:140
nots <- c(47, 127)
pots <- subset(pots, !(pots %in% nots))
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
## [19] 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
## [37] 37 38 39 40 41 42 43 44 45 46 48 49 50 51 52 53 54 55
## [55] 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
## [73] 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
## [91] 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109
## [109] 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 128
## [127] 129 130 131 132 133 134 135 136 137 138 139 140sum_clean$pot.no <- pots
sum_cleanCP04 <- rbind(CP04, sum_clean)Decoding the pot information using the FW data
We will load the data containing fresh and dry weight, as well as all of the decoding information:
decode <- read.csv("FWDW_exp4.csv")
colnames(decode)[1] <- "pot.no"
FW_data <- decode
decode <- decode[,1:3]
decode <- subset(decode, !(decode$pot.no %in% nots))
CP04 <- CP04[,c(1:3,10,4:9)]
CP04CP04_decoded <- merge(CP04, decode, all = TRUE)
unique(CP04_decoded$day)## [1] "2022.05.11" "2022.05.19" "2022.05.15" "2022.05.17" "2022.05.05"
## [6] "2022.05.07" "2022.05.13" "2022.05.09"CP04_decodedVisualize the data over time:
For this - we need to rename the date into a numeric day of experiment:
CP04_decoded$day <- gsub("2022.05.05", 0, CP04_decoded$day)
CP04_decoded$day <- gsub("2022.05.07", 2, CP04_decoded$day)
CP04_decoded$day <- gsub("2022.05.09", 4, CP04_decoded$day)
CP04_decoded$day <- gsub("2022.05.11", 6, CP04_decoded$day)
CP04_decoded$day <- gsub("2022.05.13", 8, CP04_decoded$day)
CP04_decoded$day <- gsub("2022.05.15", 10, CP04_decoded$day)
CP04_decoded$day <- gsub("2022.05.17", 12, CP04_decoded$day)
CP04_decoded$day <- gsub("2022.05.19", 14, CP04_decoded$day)
CP04_decoded$day <- as.numeric(as.character(CP04_decoded$day))
CP04_decodedNow - let’s visualize the area.sum over time for each pot, and divide colours by treatments:
unique(CP04_decoded$Treatment)## [1] "Control" "Drought" NACP04_decoded$Treatment <- gsub(" ", "", CP04_decoded$Treatment)
CP04_decoded <- na.omit(CP04_decoded)
CP04_decoded$Treatment <- factor(CP04_decoded$Treatment, levels = c("Control", "Drought"))
library(ggplot2)
library(ggpubr)
library("ggsci")
CP04_decoded$day <- as.factor(CP04_decoded$day)
Area_lgraph_CP04 <- ggplot(data=CP04_decoded, aes(x= day, y=area.sum, group = pot.no, color = Treatment))
Area_lgraph_CP04 <- Area_lgraph_CP04 + geom_line(alpha = 0.1)
Area_lgraph_CP04 <- Area_lgraph_CP04 + stat_summary(fun.data = mean_se, geom="ribbon", linetype=0, aes(group= Treatment), alpha=0.3)
Area_lgraph_CP04 <- Area_lgraph_CP04 + stat_summary(fun=mean, aes(group= Treatment), size=0.7, geom="line", linetype = "dashed")
Area_lgraph_CP04 <- Area_lgraph_CP04 + stat_compare_means(aes(group = Treatment), label = "p.signif", method = "t.test", hide.ns = T)
Area_lgraph_CP04 <- Area_lgraph_CP04 + ylab("Shoot Size (7 x SV pixels)") + xlab("Days After Stress") + scale_color_jco()
Area_lgraph_CP04
Let’s save the data too into a clean and separate csv file:
write.csv(CP04_decoded, "Cowpea_04_Clean_data.csv", row.names = FALSE)Visualize the correlation between FW and pixels at final day:
last_day <- subset(CP04_decoded, CP04_decoded$day == 14)
unique(last_day$Treatment)## [1] Control Drought
## Levels: Control Droughtunique(FW_data$Treatment)## [1] "Control" "Drought"FW_data$Treatment <- gsub(" ", "", FW_data$Treatment)
last_day_FW <- merge(last_day, FW_data, by=c("pot.no", "Genotype", "Treatment"), all=TRUE)
last_day_FWFW_Area_CP04 <- ggscatter(last_day_FW, x = "area.sum", y = "FW",rug = TRUE) + stat_cor()
FW_Area_CP04## Warning: Removed 2 rows containing non-finite values (stat_cor).## Warning: Removed 2 rows containing missing values (geom_point).