Introduction
I am located in Agriculture and Food at the Waite Campus, Adelaide. I have a relatively broad background in various fields such as fungal genetics, human biochemistry, and proteomics. In the last ten years, I was primarily involved in the use of plant molecular genetics to isolate the causal loci for apomixis (the process of asexual reproduction through seed). I am currently transitioning into grapevine research for viticulture and am involved in a project investigating genetic and mechanistic aspects of abiotic stress in grapevines. Data School has been a steep learning curve for me as I had no prior coding experience, and previous bioinformatic tasks I had directly undertaken typically involved GUI based software such as Geneious and Excel.
My Project
The dataset I used for Data School consists of chloride ion measurements in the leaves of a grapevine mapping population subjected to salt and/or heat treatments. These chloride measurements are a subset of a substantially larger experiment investigating abiotic stress which includes state of the art imaging analysis in a plant accelerator to determine growth rates and water usage. The ultimate goal is to be able to use R to develop rigourous, effective, and reproducible analyses pipelines for all aspects of the experiment. The immediate goals for this dataset in Data School was to evaluate the utility of R in analysing the chloride data subset, and to identify potential sources of variation within the data.
###Standards data for each replicate_batch
##17.1mM Cl standards data
raw_r1b1_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 9, range = "F49:J67")
#reformat standards data
#extract vectors with relevant range of values from table as vectors
a_r1b1 <- raw_r1b1_17mM_st$...3[(1:5)]
b_r1b1 <- raw_r1b1_17mM_st$`607 mg/L`[(1:5)]
c_r1b1 <- raw_r1b1_17mM_st$...3[(8:17)]
d_r1b1 <- raw_r1b1_17mM_st$`607 mg/L`[(8:17)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r1b1 <- cbind.fill(a_r1b1, b_r1b1, c_r1b1, d_r1b1, fill = NA)
#rename columns in matrix
colnames(mx_r1b1) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r1b1 <- data.frame(mx_r1b1) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r1b1 <- df_1_r1b1 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r1b1_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 9, range = "M49:O57")
#reformat bulk standards data
#rename columns in dataframe, correct bulk standard reading & calc Cl% dry weight
bulk_r1b1 <- raw_r1b1_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r1b1)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r1b1[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r1b2_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 10, range = "F48:J66")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r1b2 <- raw_r1b2_17mM_st$...3[(1:5)]
b_r1b2 <- raw_r1b2_17mM_st$`607 mg/L`[(1:5)]
c_r1b2 <- raw_r1b2_17mM_st$...3[(8:17)]
d_r1b2 <- raw_r1b2_17mM_st$`607 mg/L`[(8:17)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r1b2 <- cbind.fill(a_r1b2, b_r1b2, c_r1b2, d_r1b2, fill = NA)
#rename columns in matrix
colnames(mx_r1b2) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r1b2 <- data.frame(mx_r1b2) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r1b2 <- df_1_r1b2 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r1b2_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 10, range = "M48:O57")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r1b2 <- raw_r1b2_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r1b2)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r1b2[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r1b3_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 12, range = "B48:E66")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r1b3 <- raw_r1b3_17mM_st$...2[(1:5)]
b_r1b3 <- raw_r1b3_17mM_st$`607 mg/L`[(1:5)]
c_r1b3 <- raw_r1b3_17mM_st$...2[(8:16)]
d_r1b3 <- raw_r1b3_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r1b3 <- cbind.fill(a_r1b3, b_r1b3, c_r1b3, d_r1b3, fill = NA)
#rename columns in matrix
colnames(mx_r1b3) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r1b3 <- data.frame(mx_r1b3) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r1b3 <- df_1_r1b3 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r1b3_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 12, range = "H48:J61")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r1b3 <- raw_r1b3_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r1b3)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r1b3[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r1b4_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 14, range = "B49:E67")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r1b4 <- raw_r1b4_17mM_st$...2[(1:5)]
b_r1b4 <- raw_r1b4_17mM_st$`607 mg/L`[(1:5)]
c_r1b4 <- raw_r1b4_17mM_st$...2[(8:16)]
d_r1b4 <- raw_r1b4_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r1b4 <- cbind.fill(a_r1b4, b_r1b4, c_r1b4, d_r1b4, fill = NA)
#rename columns in matrix
colnames(mx_r1b4) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r1b4 <- data.frame(mx_r1b4) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r1b4 <- df_1_r1b4 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r1b4_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 14, range = "H49:J58")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r1b4 <- raw_r1b4_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r1b4)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r1b4[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r1b5_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 18, range = "B54:E72")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r1b5 <- raw_r1b5_17mM_st$...2[(1:5)]
b_r1b5 <- raw_r1b5_17mM_st$`607 mg/L`[(1:5)]
c_r1b5 <- raw_r1b5_17mM_st$...2[(8:17)]
d_r1b5 <- raw_r1b5_17mM_st$`607 mg/L`[(8:17)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r1b5 <- cbind.fill(a_r1b5, b_r1b5, c_r1b5, d_r1b5, fill = NA)
#rename columns in matrix
colnames(mx_r1b5) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r1b5 <- data.frame(mx_r1b5) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r1b5 <- df_1_r1b5 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r1b5_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 18, range = "H54:J65")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r1b5 <- raw_r1b5_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r1b5)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r1b5[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r1b6_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 20, range = "B38:E56")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r1b6 <- raw_r1b6_17mM_st$...2[(1:5)]
b_r1b6 <- raw_r1b6_17mM_st$`607 mg/L`[(1:5)]
c_r1b6 <- raw_r1b6_17mM_st$...2[(8:15)]
d_r1b6 <- raw_r1b6_17mM_st$`607 mg/L`[(8:15)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r1b6 <- cbind.fill(a_r1b6, b_r1b6, c_r1b6, d_r1b6, fill = NA)
#rename columns in matrix
colnames(mx_r1b6) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r1b6 <- data.frame(mx_r1b6) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r1b6 <- df_1_r1b6 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r1b6_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 20, range = "H38:J49")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r1b6 <- raw_r1b6_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r1b6)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r1b6[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r2b1_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 8, range = "F122:J140")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r2b1 <- raw_r2b1_17mM_st$...3[(1:5)]
b_r2b1 <- raw_r2b1_17mM_st$`607 mg/L`[(1:5)]
c_r2b1 <- raw_r2b1_17mM_st$...3[(8:15)]
d_r2b1 <- raw_r2b1_17mM_st$`607 mg/L`[(8:15)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r2b1 <- cbind.fill(a_r2b1, b_r2b1, c_r2b1, d_r2b1, fill = NA)
#rename columns in matrix
colnames(mx_r2b1) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r2b1 <- data.frame(mx_r2b1) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r2b1 <- df_1_r2b1 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r2b1_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 8, range = "M122:O131")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r2b1 <- raw_r2b1_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r2b1)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r2b1[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r2b2_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 9, range = "F118:J136")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r2b2 <- raw_r2b2_17mM_st$...3[(1:5)]
b_r2b2 <- raw_r2b2_17mM_st$`607 mg/L`[(1:5)]
c_r2b2 <- raw_r2b2_17mM_st$...3[(8:16)]
d_r2b2 <- raw_r2b2_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r2b2 <- cbind.fill(a_r2b2, b_r2b2, c_r2b2, d_r2b2, fill = NA)
#rename columns in matrix
colnames(mx_r2b2) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r2b2 <- data.frame(mx_r2b2) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r2b2 <- df_1_r2b2 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r2b2_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 9, range = "M118:O128")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r2b2 <- raw_r2b2_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r2b2)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r2b2[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r2b3_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 11, range = "C51:G69")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r2b3 <- raw_r2b3_17mM_st$...3[(1:5)]
b_r2b3 <- raw_r2b3_17mM_st$`607 mg/L`[(1:5)]
c_r2b3 <- raw_r2b3_17mM_st$...3[(8:16)]
d_r2b3 <- raw_r2b3_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r2b3 <- cbind.fill(a_r2b3, b_r2b3, c_r2b3, d_r2b3, fill = NA)
#rename columns in matrix
colnames(mx_r2b3) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r2b3 <- data.frame(mx_r2b3) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r2b3 <- df_1_r2b3 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r2b3_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 11, range = "J51:L61")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r2b3 <- raw_r2b3_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r2b3)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r2b3[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r2b4_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 13, range = "C74:G92")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r2b4 <- raw_r2b4_17mM_st$...3[(1:5)]
b_r2b4 <- raw_r2b4_17mM_st$`607 mg/L`[(1:5)]
c_r2b4 <- raw_r2b4_17mM_st$...3[(8:16)]
d_r2b4 <- raw_r2b4_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r2b4 <- cbind.fill(a_r2b4, b_r2b4, c_r2b4, d_r2b4, fill = NA)
#rename columns in matrix
colnames(mx_r2b4) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r2b4 <- data.frame(mx_r2b4) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r2b4 <- df_1_r2b4 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r2b4_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 13, range = "J74:L85")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r2b4 <- raw_r2b4_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r2b4)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r2b4[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r2b5_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 17, n_max = 47, range = "A52:E68")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r2b5 <- raw_r2b4_17mM_st$...3[(1:5)]
b_r2b5 <- raw_r2b4_17mM_st$`607 mg/L`[(1:5)]
c_r2b5 <- raw_r2b4_17mM_st$...3[(8:16)]
d_r2b5 <- raw_r2b4_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r2b5 <- cbind.fill(a_r2b5, b_r2b5, c_r2b5, d_r2b5, fill = NA)
#rename columns in matrix
colnames(mx_r2b5) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r2b5 <- data.frame(mx_r2b5) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r2b5 <- df_1_r2b5 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r2b5_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 17, n_max = 47, range = "H52:J68")#note: no measurement data available
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r2b5 <- raw_r2b5_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r2b5)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r2b5[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
####No measured bulk standard data for rep_batch r2b5
##17.1mM Cl standards data
raw_r2b6_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 19, n_max = 53, range = "A57:E74")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r2b6 <- raw_r2b6_17mM_st$...3[(1:5)]
b_r2b6 <- raw_r2b6_17mM_st$`607 mg/L`[(1:5)]
c_r2b6 <- raw_r2b6_17mM_st$...3[(8:16)]
d_r2b6 <- raw_r2b6_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r2b6 <- cbind.fill(a_r2b6, b_r2b6, c_r2b6, d_r2b6, fill = NA)
#rename columns in matrix
colnames(mx_r2b6) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r2b6 <- data.frame(mx_r2b6) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r2b6 <- df_1_r2b6 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r2b6_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 19, n_max = 53, range = "H57:J68")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r2b6 <- raw_r2b6_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r2b6)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r2b6[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r3b1_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 8, n_max = 44, range = "F50:J65")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r3b1 <- raw_r3b1_17mM_st$...3[(1:5)]
b_r3b1 <- raw_r3b1_17mM_st$`607 mg/L`[(1:5)]
c_r3b1 <- raw_r3b1_17mM_st$...3[(8:15)]
d_r3b1 <- raw_r3b1_17mM_st$`607 mg/L`[(8:15)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r3b1 <- cbind.fill(a_r3b1, b_r3b1, c_r3b1, d_r3b1, fill = NA)
#rename columns in matrix
colnames(mx_r3b1) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r3b1 <- data.frame(mx_r3b1) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r3b1 <- df_1_r3b1 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r3b1_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 8, n_max = 44, range = "M50:O59")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r3b1 <- raw_r3b1_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r3b1)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r3b1[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r3b2_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 9, n_max = 43, range = "F48:J62")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r3b2 <- raw_r3b2_17mM_st$...3[(1:5)]
b_r3b2 <- raw_r3b2_17mM_st$`607 mg/L`[(1:5)]
c_r3b2 <- raw_r3b2_17mM_st$...3[(8:13)]
d_r3b2 <- raw_r3b2_17mM_st$`607 mg/L`[(8:13)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r3b2 <- cbind.fill(a_r3b2, b_r3b2, c_r3b2, d_r3b2, fill = NA)
#rename columns in matrix
colnames(mx_r3b2) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r3b2 <- data.frame(mx_r3b2) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r3b2 <- df_1_r3b2 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r3b2_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 9, n_max = 43, range = "M48:O58")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r3b2 <- raw_r3b2_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r3b2)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r3b2[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r3b3_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 11, n_max = 44, range = "C48:G64")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r3b3 <- raw_r3b3_17mM_st$...3[(1:5)]
b_r3b3 <- raw_r3b3_17mM_st$`607 mg/L`[(1:5)]
c_r3b3 <- raw_r3b3_17mM_st$...3[(8:16)]
d_r3b3 <- raw_r3b3_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r3b3 <- cbind.fill(a_r3b3, b_r3b3, c_r3b3, d_r3b3, fill = NA)
#rename columns in matrix
colnames(mx_r3b3) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r3b3 <- data.frame(mx_r3b3) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r3b3 <- df_1_r3b3 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r3b3_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 11, n_max = 44, range = "K48:M58")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r3b3 <- raw_r3b3_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r3b3)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r3b3[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r3b4_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 13, n_max = 48, range = "C53:G69")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r3b4 <- raw_r3b4_17mM_st$...3[(1:5)]
b_r3b4 <- raw_r3b4_17mM_st$`607 mg/L`[(1:5)]
c_r3b4 <- raw_r3b4_17mM_st$...3[(8:16)]
d_r3b4 <- raw_r3b4_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r3b4 <- cbind.fill(a_r3b4, b_r3b4, c_r3b4, d_r3b4, fill = NA)
#rename columns in matrix
colnames(mx_r3b4) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r3b4 <- data.frame(mx_r3b4) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r3b4 <- df_1_r3b4 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r3b4_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 13, n_max = 48, range = "J53:L63")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r3b4 <- raw_r3b4_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r3b4)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r3b4[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##17.1mM Cl standards data
raw_r3b5_17mM_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 17, n_max = 45, range = "B54:E69")
#reformat standards data
#extract vectors with relevant range of values from table as vectors and convert to numeric
a_r3b5 <- raw_r3b5_17mM_st$...2[(1:5)]
b_r3b5 <- raw_r3b5_17mM_st$`607 mg/L`[(1:5)]
c_r3b5 <- raw_r3b5_17mM_st$...2[(8:15)]
d_r3b5 <- raw_r3b5_17mM_st$`607 mg/L`[(8:16)]
#combine vectors to create a matrix and fill in missing values with NA. Done as vectors are of different length
mx_r3b5 <- cbind.fill(a_r3b5, b_r3b5, c_r3b5, d_r3b5, fill = NA)
#rename columns in matrix
colnames(mx_r3b5) <- c("preassay_blank", "preassay_17mM_standard", "intra_assay_blank", "intra_assay_17mM_standard")
#convert matrix to a data frame. Note: column naming problem if I try to use tibble()
#subtract blanks from standard values, tidy and filter for adjusted data
df_1_r3b5 <- data.frame(mx_r3b5) %>%
mutate("preassay_standard_adj" = preassay_17mM_standard - preassay_blank,
"intra_assay_standard_adj" = intra_assay_17mM_standard - intra_assay_blank) %>%
gather(key = "standard", value = "reading_adj", na.rm=TRUE) %>%
filter(str_detect(standard, "preassay_standard_adj|intra_assay_standard_adj"))
#calculate conversion factor for Cl% dry wt calc in chloride_data and bulk_standards
cf_r3b5 <- df_1_r3b5 %>%
filter(standard == "intra_assay_standard_adj") %>%
summarise(conv_factor = 17.1 / mean(reading_adj))
###bulk grapevine data
raw_r3b5_grape_st <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 17, n_max = 45, range = "H54:J65")
#reformat bulk standards data
#rename columns in dataframe, adjust bulk standard reading & calc Cl% dry weight
bulk_r3b5 <- raw_r3b5_grape_st %>%
dplyr::rename(intra_assay_bulk_vial = `...1`, intra_assay_bulk_weight_mg = wt., intra_assay_bulk_standard_read = read) %>%
mutate("intra_assay_bulk_st_adj" = intra_assay_bulk_standard_read - mean(c_r3b5)) %>%
mutate("Cl_%_dry_weight" = ((intra_assay_bulk_st_adj * cf_r3b5[1,1]) * 35.5)/(intra_assay_bulk_weight_mg * 10)) %>%
filter(intra_assay_bulk_st_adj !=is.na(intra_assay_bulk_st_adj))
##Note: r3b5 is last rep_batch for this experiment
####Process chloride data
#assign primary data files to variables
#files with chloride data for each replicate & batch
#import rows from specific sheets with chloride data for samples only (not standards)
#Biological replicate #1, sample batches #1 - #6, rows with pertinent data
raw_r1b1 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 9, n_max = 45) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(2, 3, 4, 5, 18, 19, 22, 23, 26, 27, 28)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_10, smarthouse = smarthouse_2, cl_rep1_reading = cl_r1_read, cl_rep_2_reading = cl_r2_read,
vial_number_1 = chloride_number_29, vial_number_2 = chloride_number_30) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r1b1), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r1b1)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r1b1[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r1b1[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r1b1$rep_batch="r1b1" #add assay batch info
raw_r1b2 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 10, n_max = 44) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(2, 3, 4, 5, 18, 19, 22, 23, 26, 27, 28)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_10, smarthouse = smarthouse_2,
cl_rep1_reading = cl_r1_read, cl_rep_2_reading = cl_r2_read) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r1b2), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r1b2)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r1b2[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r1b2[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r1b2$rep_batch="r1b2" #add assay batch info
raw_r1b3 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 12, n_max = 43) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(14, 15, 18, 19, 22, 23, 24)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_6, cl_rep1_reading = cl_r1_read, cl_rep_2_reading = cl_r2_read) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r1b3), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r1b3)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r1b3[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r1b3[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r1b3$rep_batch="r1b3" #add assay batch info
raw_r1b4 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 14, n_max = 41) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(14, 15, 18, 19, 22, 23, 24)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_6, cl_rep1_reading = cl_r1_read, cl_rep_2_reading = cl_r2_read) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r1b4), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r1b4)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r1b4[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r1b4[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r1b4$rep_batch="r1b4" #add assay batch info
raw_r1b5 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 18, n_max = 48) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(16, 17, 20, 21, 22)) %>% #remove columns with analyses or duplicated info
dplyr::rename(cl_rep1_reading = cl_r1_read, cl_rep_2_reading = cl_r2_read) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r1b5), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r1b5)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r1b5[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r1b5[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r1b5$rep_batch="r1b5" #add assay batch info
raw_r1b6 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP1.xlsx", sheet = 20, n_max = 30) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(16, 17, 20, 21, 22)) %>% #remove columns with analyses or duplicated info
dplyr::rename(cl_rep1_reading = cl_r1_read, cl_rep_2_reading = cl_r2_read) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r1b6), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r1b6)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r1b6[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r1b6[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r1b6$rep_batch="r1b6"#add assay batch info
#Biological replicate #2, sample batches #1 - #6
raw_r2b1 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 8, cell_rows(73:115)) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(2, 3, 4, 5, 18, 19, 23, 24, 28, 29, 30)) %>% #remove columns with analyses or duplicated info
dplyr::rename(smarthouse = smarthouse_2, count_number = count) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r2b1), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r2b1)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r2b1[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r2b1[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r2b1$rep_batch="r2b1" #add assay batch info
raw_r2b2 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 9, cell_rows(71:114)) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(2, 3, 4, 5, 18, 19, 23, 24, 28, 29, 30)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_10, smarthouse = smarthouse_2) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r2b2), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r2b2)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r2b2[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r2b2[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r2b2$rep_batch="r2b2" #add assay batch info
raw_r2b3 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 11, n_max = 47) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(14, 15, 19, 20, 24, 25, 26)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_6) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r2b3), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r2b3)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r2b3[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r2b3[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r2b3$rep_batch="r2b3" #add assay batch info
raw_r2b4 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 13, cell_rows(29:69)) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(14, 15, 19, 20, 24, 25, 26)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_6) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r2b4), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r2b4)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r2b4[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r2b4[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r2b4$rep_batch="r2b4" #add assay batch info
raw_r2b5 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 17, cell_rows(74:120)) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(14, 15)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_6) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r2b5), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r2b5)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r2b5[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r2b5[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r2b5$rep_batch="r2b5" #add assay batch info
raw_r2b6 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP2.xlsx", sheet = 19, n_max = 53) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(14, 15, 19, 20, 24, 25, 26)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_6) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r2b5), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r2b5)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r2b5[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r2b5[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r2b6$rep_batch="r2b6" #add assay batch info
#Biological replicate #3, sample batches #1 - #6
raw_r3b1 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 8, n_max = 44) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(2, 3, 4, 5, 18, 19, 23, 24, 28, 29, 30)) %>% #remove columns with analyses or duplicated info
dplyr::rename(smarthouse = smarthouse_2, count_number = count) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r3b1), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r3b1)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r3b1[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r3b1[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r3b1$rep_batch="r3b1" #add assay batch info
raw_r3b2 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 9, n_max = 43) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(2, 3, 4, 5, 18, 19, 23, 24, 28, 29, 30)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_10, smarthouse = smarthouse_2) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r3b2), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r3b2)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r3b2[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r3b2[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r3b2$rep_batch="r3b2" #add assay batch info
raw_r3b3 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 11, n_max = 44) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(4, 15, 16, 20, 21, 25, 26, 27)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_7) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r3b3), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r3b3)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r3b3[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r3b3[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r3b3$rep_batch="r3b3" #add assay batch info
raw_r3b4 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 13, n_max = 48) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(14, 15, 19, 20, 24, 25, 26)) %>% #remove columns with analyses or duplicated info
dplyr::rename(rep = rep_6) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r3b4), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r3b4)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r3b4[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r3b4[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r3b4$rep_batch="r3b4" #add assay batch info
raw_r3b5 <- read_xlsx("Data/3&4. 0192 barcodes (LAM)-BIOMASS.REP3.DB_SH.xlsx", sheet = 17, n_max = 45) %>%
clean_names() %>% #replaces spaces and removes symbols from column names
select(-c(16, 17, 20, 21, 22)) %>% #remove columns with analyses or duplicated info
dplyr::rename(cl_rep1_reading = cl_r1_read, cl_rep_2_reading = cl_r2_read) %>% #rename columns with variant names, need to force through dplyr because of plyr issue
mutate("cl_rep1_reading_adj" = as.numeric(cl_rep1_reading) - mean(c_r3b5), "cl_rep2_reading_adj" = as.numeric(cl_rep_2_reading) - mean(c_r3b5)) %>%
mutate("cl_%_dry_weight_rep1" = (cl_rep1_reading_adj * cf_r3b5[1,1] * 35.5) / (as.numeric(cl_rep1_wt_mg) * 10),
"cl_%_dry_weight_rep2" = (cl_rep2_reading_adj * cf_r3b5[1,1] * 35.5) / (as.numeric(cl_rep2_wt_mg) * 10))
raw_r3b5$rep_batch="r3b5" #add assay batch info
#joining data together
dfs <- list(raw_r1b1, raw_r1b2, raw_r1b3, raw_r1b4, raw_r1b5, raw_r1b6,
raw_r2b1, raw_r2b2, raw_r2b3, raw_r2b4, raw_r2b5, raw_r2b6,
raw_r3b1, raw_r3b2, raw_r3b3, raw_r3b4, raw_r3b5)#list of dataframes with raw data
chloride_dfs <- join_all(dfs, type = "full") #join dataframes together
#renaming columns in combined dataframe
chloride_1_data <- chloride_dfs %>%
dplyr::rename(bio_rep = rep, treatment_heat = treat_1_h, treatment_salt = treat_2_s,
cl_sample_weight_mg_rep1 = cl_rep1_wt_mg, cl_read_rep1 = cl_rep1_reading,
cl_sample_weight_mg_rep2 = cl_rep2_wt_mg, cl_read_rep2 = cl_rep_2_reading,
cl_read_adj_rep1 = cl_rep1_reading_adj, cl_read_adj_rep2 = cl_rep2_reading_adj,
cl_dry_weight_calc_rep1 = "cl_%_dry_weight_rep1", cl_dry_weight_calc_rep2 = "cl_%_dry_weight_rep2",
vial_number_rep1 = vial_number_1, vial_number_rep2 = vial_number_2) #rename columns with variant names, need to force through dplyr because of plyr issue
#changing treatment value names
chloride_1_data$treatment_salt[is.na(chloride_1_data$treatment_salt)] <- "no_salt"
chloride_1_data$treatment_salt[str_detect(chloride_1_data$treatment_salt, "Salt")] <- "salt"
chloride_1_data$treatment_heat[str_detect(chloride_1_data$treatment_heat, "CONT.")] <- "no_heat"
chloride_1_data$treatment_heat[str_detect(chloride_1_data$treatment_heat, "HEAT")] <- "heat"
#changing some renamed columns to numerical data
chloride_1_data$cl_sample_weight_mg_rep1 <- as.numeric(as.character(chloride_1_data$cl_sample_weight_mg_rep1))
chloride_1_data$cl_sample_weight_mg_rep2 <- as.numeric(as.character(chloride_1_data$cl_sample_weight_mg_rep2))
chloride_1_data$cl_read_rep1 <- as.numeric(as.character(chloride_1_data$cl_read_rep1))
chloride_1_data$cl_read_rep2 <- as.numeric(as.character(chloride_1_data$cl_read_rep2))
chloride_1_data$bio_rep <- as.character((chloride_1_data$bio_rep))
#add harvester data
raw_harvester <- read_xlsx("Data/0192 Mature laminae chloride summary with harvester info.xlsx") %>%
clean_names() %>%
select(code, harvester)
chloride_data_harv <- left_join(chloride_1_data, raw_harvester, by="code")
#add laminae sampling weight data
raw_laminae <- read_xlsx("Data/0192 Mature Laminae dry wts.xlsx") %>%
clean_names() %>%
dplyr::rename(laminae_sample_weight = dry_wt_lam_ion) %>%
select(code, laminae_sample_weight)
chloride_data_harv_lam <- left_join(chloride_data_harv, raw_laminae, by = "code")
chloride_data_harv_lam$laminae_sample_weight <- as.numeric(as.character(chloride_data_harv_lam$laminae_sample_weight))
#remove columns no longer required
chloride_data_harv_lam1 <- chloride_data_harv_lam %>%
select(-c(vial_number_rep1, vial_number_rep2, cl_read_rep1, cl_read_rep2, cl_read_adj_rep1, cl_read_adj_rep2))
#using gather, separate and spread
chloride_1_tidy <- chloride_data_harv_lam1 %>%
#gather both variables/duplicates into key and value columns
gather(key = assay_rep, value = value, "cl_sample_weight_mg_rep1", "cl_sample_weight_mg_rep2", "cl_dry_weight_calc_rep1", "cl_dry_weight_calc_rep2") %>%
#separate the gathered column variable into two columns, one with the assay info and the other with replicate info
separate(col = assay_rep, into = c("assay", "tech_rep"), sep = "_rep") %>%
#spread assay column into two columns with assay info
spread(key = assay, value = value)
#cleanup to remove NA values in sample weight and cl dry weight columns
chloride_2_tidy <- chloride_1_tidy %>%
# mutate(tech_rep = as.numeric(tech_rep)) %>%
filter(!is.na(cl_sample_weight_mg) & !is.na(cl_dry_weight_calc))
#remove outlier in cl_sample_weight_mg tech_rep1.
chloride_3_tidy <- chloride_2_tidy %>%
filter(cl_sample_weight_mg <70)
#combine treatments into a single column
chloride_4_tidy <- chloride_3_tidy %>%
unite(col = "treatment", "treatment_heat", "treatment_salt", sep = "_&_")
#Normality tests
#assign objects to 3 variables
laminae_sample_weight <- chloride_3_tidy$laminae_sample_weight
cl_sample_weight_mg <- chloride_3_tidy$cl_sample_weight_mg
cl_dry_weight_calc <- chloride_3_tidy$cl_dry_weight_calcPreliminary results
In this experiment, a grapevine population was subjected to the following salt and heat treatments to assess the effects on chloride uptake into the leaves:
- no salt, no heat (control)
- salt, no heat
- no salt, heat
- salt, heat
Three biological replicates of each grapevine were assessed in each treatment group and the chloride measurement was undertaken in duplicate.
Initial analyses were focussed on assessing the distribution and robustness of the data as follows:
First, an assessment of the distribution of the chloride measurements in each treatment group indicated there was a positive skew in these data. This is shown in the representative histogram of the leaf chloride levels in the control group (Figure 1). QQ plots of the calculated chloride data further supported the notion that the data was not normally distributed (representative plot of the control group shown in Figure 2. Four different statistical tests (Shapiro-Wilk; Anderson-Darling; Kolmogorov-Smirnov; Cramer-von Mises) were performed and further confirmed that the data was highly unlikely to be normal (data not shown). Together, these anayses indicates that the data is not normally distributed and non-parametric statistical tests will be required for future statistical analysis of this dataset.
Second, the consistency of the three biological data replicates was visually assessed for each treatment group (Figure 3). The median and data distribution for the three biological replicates was largely consistent with the exception of the heat & salt treatment group where replicate 1 had a higher median than the other two biological replicates, the significance and any potential ramifications of this difference will be examined further.
Third, as the harvesting of leaf samples was undertaken by multiple people, the data generated by each harvester were visualised to determine if there were any notable variations (Figure 4). Overall the data distrubtions looked relatively consistent between harvesters. Some variation was expected due to genotypic differences in the grapevine lines analysed.
Based on the preliminary analysis above, the data appears to be positively skewed but reasonably robust.
In terms of the experimental results, the analyses clearly shows that at basal salt levels, heat has little effect on salt uptake. However, when exposed to high salinity, heat has a substantial effect on the uptake of chloride ions into grapevine leaves (Figure 5).
Furthermore, there is substantial variation between the individual grapevine lines in the uptake of chloride in the presence of high salinity with and without heat (Figure 6.) and likely reflects segregation of genetic factors within the grapevine mapping population. This will be a particular point of interest in future analyses.
nhns <- chloride_3_tidy %>%
filter(treatment_heat == "no_heat", treatment_salt == "no_salt")
nhns1 <- nhns$cl_dry_weight_calc
hist(nhns1, main = " ", xlab = "Leaf chloride content")
Figure 1: Distribution of leaf chloride measurements in control treatment group
Figure 2: QQ Norm plot of leaf chloride content
#check biological reps
plot_0b <- chloride_4_tidy %>%
ggplot(aes(x=bio_rep, y=cl_dry_weight_calc, group = bio_rep, colour = treatment, shape = tech_rep)) +
geom_jitter(size = 2, alpha =0.5)+
theme(axis.text.x=element_text(size=15, angle=0), axis.title=element_text(size=15), axis.text.y=element_text(size=15), legend.title = element_text(size=15), legend.text = element_text(size=15), legend.position = "bottom")+
guides(colour=guide_legend(override.aes=list(size=3)))+
ylab("Leaf chloride content")+
xlab("Biological replicate")+
stat_summary(
geom = "point",
fun.y = "median",
col ="black",
size = 3,
shape = 23, fill = "brown",
)+
facet_wrap(~treatment)+
theme(strip.text.x = element_text(size = 15))
plot_0b
Figure 3: Comparison of biological reps
#Check harvester effects - as per 1a but harvester separated
plot_2a <- chloride_3_tidy %>%
unite(col = "treatment", "treatment_heat", "treatment_salt", sep = "_&_") %>%
ggplot(aes(x=harvester, y=cl_dry_weight_calc, colour = treatment)) +
geom_jitter(size = 2, alpha =0.5)+
theme(axis.text.x=element_text(size=15, angle=45, hjust=1), axis.title=element_text(size=15), axis.text.y=element_text(size=15), legend.title = element_text(size=15), legend.text = element_text(size=15), legend.position = "bottom")+
guides(colour=guide_legend(override.aes=list(size=3)))+
ylab("Leaf chloride content")+
xlab("Harvester")+
stat_summary(
geom = "point",
fun.y = "median",
col ="black",
size = 3,
shape = 23, fill = "brown",
)+
facet_grid(~treatment, scales = "fixed")+
theme(strip.text.x = element_text(size = 15))
plot_2a
Figure 4: leaf chloride content in samples harvested by different people
#compare effects of heat and salt treatments
#combined treatments into a single column and combined bio_reps
plot_1a <- chloride_3_tidy %>%
unite(col = "treatment", "treatment_heat", "treatment_salt", sep = "_&_") %>%
ggplot(aes(x=treatment, y=cl_dry_weight_calc, colour = treatment)) +
geom_jitter(size = 2, alpha =0.5)+
theme(axis.text.x=element_text(size=15, angle=45, hjust=1), axis.title=element_text(size=15), axis.text.y=element_text(size=15), legend.title = element_text(size=15), legend.text = element_text(size=15), legend.position = "bottom")+
guides(colour=guide_legend(override.aes=list(size=3)))+
ylab("Leaf chloride content")+
xlab("Treatment")+
stat_summary(
geom = "point",
fun.y = "median",
col ="black",
size = 3,
shape = 23, fill = "brown",
)
plot_1a
Figure 5: Leaf chloride content in grapevine with heat &/or salt treatment
#compare effects of heat and salt treatments
#combined treatments into a single column and combined bio_reps
plot_100 <- chloride_4_tidy %>%
ggplot(aes(x=genotype_id, y=cl_dry_weight_calc, colour = treatment))+
xlab("Grapevine lines")+
ylab("Leaf chloride content")+
geom_boxplot(size=0.5)+
theme(axis.text.x = element_text(angle=90, hjust =1), legend.title = element_text(size=15), legend.text = element_text(size=15), legend.position = "bottom")
plot_100
Figure 6: Leaf chloride content in each grapevine line
My Digital Toolbox
It quickly became apparent to me that utilising R effectively requires a flexible mindset in trying and using tools available in R packages outside of the tidyverse. In my analyses, I used tools available in the following packages:
- ggplot2
- readxl
- plyr
- dplyr
- janitor
- rowr
- nortest
- pca3d
- ggfortify
Favourite tool (optional)
Github: This was particularly useful as I was developing the code with three different computers at various times.
R Markdown: This is a fantastic tool to efficiently communicate results to others and fits seamlessly with the version control properties of Git.
My time went …
The most time consuming aspect BY FAR was wrangling the data. The dataset I have used was provided by my colleagues and was spread across ~20 different spreadsheets with inconsistent column locations and headers. In addition, the spreadsheets contained raw and analysed data, redundant data, embedded data in different formats etc. In other words, typical data spreadsheets used by non-programmers!
I wrote ~770 lines of code to import, arrange and ‘tidy’ these data into a suitable format before I could commence analyses in R.
The most surprising and frustrating challenge I encountered was my inability to effectively use the in-built help files in R, and peer-support provided in R-related websites (concepts, code, and examples), as the information provided was typically at a level beyond what I could readily understand and implement.
The online material readily accessible to beginneRs always seem to use the same inbuilt ‘perfect’ datasets, so issues associated with ‘real-world’ datasets are typically not addressed and can be extremely time consuming to overcome.
Next steps
Data School: Focus has provided an excellent foundation in data analyses using R, and in data management in general. Clearly, there is much still to be learned and I am keen to continue to apply and expand on these newly acquired skills in my future project work. I will be expanding the preliminary analyses presented here to the complete experimental dataset and work with the Project Leader and team members towards implementing R analyses pipelines in future experiments.
My Data School Experience
Overall, Data School: Focus has been an excellent experience. The most useful and enjoyable aspects of the course were the practical days where the knowledge learned could be implemented and reinforced with our own datasets and at our own pace, with readily accessible support to help overcome difficulties and issues as they arose.
In terms of transferring knowledge to my colleagues, I have had informal discussions with Project Leaders and team members about the potential benefits of applying R and in adopting best practices for data recording, storage, and formatting.