Cowpea Screen 01
This notebook describes the data analysis pipeline for side-view images collected for 1st cowpea screen (Cowpea 01), performend between February 11th and March 15th, 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] 7data <- 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
“Pot 18 (524-B, drought) did not germinate
Pot 92 and 97 (TVu-14346, control and drought) did not germinate
Pot 113 (TVu-16237, control) did not germinate; pot 118 (TVu-16237, drought) germinated late
Pot 137 (TVu-6643, drought) germinated on March 12
Pot 2 (CB5-2, control) on before watering on March 15th is missing
Look at images of: 123+128, 124+129. They all have bumps on the leaves”
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
nots <- c(18, 92, 97, 113, 137)
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 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] 93 94 95 96 98 99 100 101 102 103 104 105 106 107 108 109 110 111
## [109] 112 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
## [127] 131 132 133 134 135 136 138 139 140sum_clean$pot.no <- pots
sum_cleanCP01 <- sum_cleanOK - now let’s do next day but much quicker:
data <- read.csv(my_files[2])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$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$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanpots <- 1:140
nots <- c(18, 92, 97, 113, 137)
pots <- subset(pots, !(pots %in% nots))
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 19
## [19] 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
## [37] 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
## [55] 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
## [73] 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
## [91] 93 94 95 96 98 99 100 101 102 103 104 105 106 107 108 109 110 111
## [109] 112 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
## [127] 131 132 133 134 135 136 138 139 140sum_clean$pot.no <- pots
sum_cleanCP01 <- rbind(CP01, 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_cleanPot 126 imaged twice
sum_clean <- subset(sum_clean, sum_clean$time != "15.33.45")
pots <- 1:140
nots <- c(18, 92, 97, 113, 137)
pots <- subset(pots, !(pots %in% nots))
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 19
## [19] 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
## [37] 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
## [55] 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
## [73] 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
## [91] 93 94 95 96 98 99 100 101 102 103 104 105 106 107 108 109 110 111
## [109] 112 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
## [127] 131 132 133 134 135 136 138 139 140sum_clean$pot.no <- pots
sum_cleanCP01 <- rbind(CP01, 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_cleanadditionally missing pot 103 and 119
pots <- 1:140
nots <- c(18, 92, 97, 103, 113, 119, 137)
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 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] 93 94 95 96 98 99 100 101 102 104 105 106 107 108 109 110 111 112
## [109] 114 115 116 117 118 120 121 122 123 124 125 126 127 128 129 130 131 132
## [127] 133 134 135 136 138 139 140sum_clean$pot.no <- pots
sum_cleanCP01 <- rbind(CP01, sum_clean)data <- read.csv(my_files[5])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$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$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanpots <- 1:140
nots <- c(18, 27, 57, 78, 92, 97, 113, 137)
pots <- subset(pots, !(pots %in% nots))
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 19
## [19] 20 21 22 23 24 25 26 28 29 30 31 32 33 34 35 36 37 38
## [37] 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
## [55] 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
## [73] 76 77 79 80 81 82 83 84 85 86 87 88 89 90 91 93 94 95
## [91] 96 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 114 115
## [109] 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133
## [127] 134 135 136 138 139 140sum_clean$pot.no <- pots
sum_cleanCP01 <- rbind(CP01, 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_cleanpots <- 1:140
nots <- c(18, 92, 97, 113)
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 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] 93 94 95 96 98 99 100 101 102 103 104 105 106 107 108 109 110 111
## [109] 112 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
## [127] 131 132 133 134 135 136 137 138 139 140sum_clean$pot.no <- pots
sum_cleanCP01 <- rbind(CP01, sum_clean)Started to image pot 137 from this timepoint!
data <- read.csv(my_files[7])
my_data <- data[,c(31, 18:25,27:30)]
for(i in 1:nrow(my_data)){
my_data$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$timestamp[i], "-")[[1]][2]
}
sum_clean <- sum_clean[,c(1,8:9,2:7)]
sum_cleanpots <- 1:140
nots <- c(18, 92, 97, 113)
pots <- subset(pots, !(pots %in% nots))
pots## [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 19
## [19] 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
## [37] 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
## [55] 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
## [73] 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91
## [91] 93 94 95 96 98 99 100 101 102 103 104 105 106 107 108 109 110 111
## [109] 112 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
## [127] 131 132 133 134 135 136 137 138 139 140sum_clean$pot.no <- pots
sum_cleanCP01 <- rbind(CP01, 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_exp1.csv")
colnames(decode)[1] <- "pot.no"
FW_data <- decode
decode <- decode[,1:3]
decode <- subset(decode, !(decode$pot.no %in% nots))
CP01 <- CP01[,c(1:3,10,4:9)]
CP01CP01_decoded <- merge(CP01, decode, all = TRUE)
unique(CP01_decoded$day)## [1] "2022.03.03" "2022.03.05" "2022.03.10" "2022.03.08" "2022.03.14"
## [6] "2022.03.17" "2022.03.12" "2022.03.04" NACP01_decodedVisualize the data over time:
For this - we need to rename the date into a numeric day of experiment:
CP01_decoded$day <- gsub("2022.03.03", 0, CP01_decoded$day)
CP01_decoded$day <- gsub("2022.03.04", 1, CP01_decoded$day)
CP01_decoded$day <- gsub("2022.03.05", 2, CP01_decoded$day)
CP01_decoded$day <- gsub("2022.03.08", 5, CP01_decoded$day)
CP01_decoded$day <- gsub("2022.03.10", 7, CP01_decoded$day)
CP01_decoded$day <- gsub("2022.03.12", 9, CP01_decoded$day)
CP01_decoded$day <- gsub("2022.03.14", 11, CP01_decoded$day)
CP01_decoded$day <- gsub("2022.03.17", 14, CP01_decoded$day)
CP01_decoded$day <- as.numeric(as.character(CP01_decoded$day))
CP01_decodedNow - let’s visualize the area.sum over time for each pot, and divide colours by treatments:
unique(CP01_decoded$Treatment)## [1] "Control " "Drought" "Control"CP01_decoded$Treatment <- gsub(" ", "", CP01_decoded$Treatment)
CP01_decoded$Treatment <- factor(CP01_decoded$Treatment, levels = c("Control", "Drought"))
library(ggplot2)
library(ggpubr)
library("ggsci")
CP01_decoded$day <- as.factor(CP01_decoded$day)
Area_lgraph_CP01 <- ggplot(data=CP01_decoded, aes(x= day, y=area.sum, group = pot.no, color = Treatment))
Area_lgraph_CP01 <- Area_lgraph_CP01 + geom_line(alpha = 0.1)
Area_lgraph_CP01 <- Area_lgraph_CP01 + stat_summary(fun.data = mean_se, geom="ribbon", linetype=0, aes(group= Treatment), alpha=0.3)
Area_lgraph_CP01 <- Area_lgraph_CP01 + stat_summary(fun=mean, aes(group= Treatment), size=0.7, geom="line", linetype = "dashed")
Area_lgraph_CP01 <- Area_lgraph_CP01 + stat_compare_means(aes(group = Treatment), label = "p.signif", method = "t.test", hide.ns = T)
Area_lgraph_CP01 <- Area_lgraph_CP01 + ylab("Shoot Size (7 x SV pixels)") + xlab("Days After Stress") + scale_color_jco()
Area_lgraph_CP01## Warning: Removed 2 rows containing non-finite values (stat_summary).
## Removed 2 rows containing non-finite values (stat_summary).## Warning: Removed 2 rows containing non-finite values (stat_compare_means).## Warning: Removed 2 row(s) containing missing values (geom_path).
Let’s save the data too into a clean and separate csv file:
write.csv(CP01_decoded, "Cowpea_01_Clean_data.csv", row.names = FALSE)Visualize the correlation between FW and pixels at final day:
last_day <- subset(CP01_decoded, CP01_decoded$day == 14)
unique(last_day$Treatment)## [1] Control Drought
## Levels: Control Droughtunique(FW_data$Treatment)## [1] "Control " "Drought" "Control"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_FW$FW <- as.numeric(as.character(last_day_FW$FW))## Warning: NAs introduced by coercionFW_Area_CP01 <- ggscatter(last_day_FW, x = "area.sum", y = "FW",rug = TRUE) + stat_cor()
FW_Area_CP01## Warning: Removed 6 rows containing non-finite values (stat_cor).## Warning: Removed 6 rows containing missing values (geom_point).