Script-4

# git pull
# or git clone if you dont have this in your pc
#also is the same of folder selector ( up righ corner), and do new folder from git
# edit your script
# git add .
# git commit -m "update script"
# git push

#first sync the fork in CytometryInR folder
# the bring all local in the positron by pull request
# and then do the new folder project and copy the data
#and then do the first commit 
#and then do the fist push to get the data in github
# for homeworks will be branch to branch transfer, more details in week 5

# the complete guide: https://umgcccfcsr.github.io/CytometryInR/course/04_IntroToTidyverse/
#the youtube video: https://www.youtube.com/watch?v=luC7SY4RJcU

#getwd() to see in which file path you are.
#setwd to change in the path file. 

#to work with FCS files and exel sheets in a 
# easier maner than copied and pasted contents of columns
# changing names, etc
#so to tidy the data
# use of the tidyverse package
#The dataset we will be using today is a manually-gated spectral flow cytometry dataset
#from proliferal blood
# also you can use your own dataset using the data.frame by extracting the matrix.
#function from the utils package (included in our base R software installation) to read it into R
# so we do not need to load a package

thefilepath <- file.path("data","Dataset.csv")
#you can specify the name itself and the file type
#you bypass the list.files step
#thefilepath is more robust

setwd("C:/Users/al2271/OneDrive - University of Cambridge/Documents/Anjana/Knowledge/R/Week-4/04_IntroToTidyverse")
setwd("C:/Users/anjan/Downloads/R course/Week-4/04_IntroToTidyverse") #home 

#functions for exel files read.csv is already in the main packages
data <- read.csv(file=thefilepath, check.names=FALSE)

colnames(data)
#it retrieve all colum names that are present in the datase
 [1] "bid"               "timepoint"         "Condition"        
 [4] "Date"              "infant_sex"        "ptype"            
 [7] "root"              "singletsFSC"       "singletsSSC"      
[10] "singletsSSCB"      "CD45"              "NotMonocytes"     
[13] "nonDebris"         "lymphocytes"       "live"             
[16] "Dump+"             "Dump-"             "Tcells"           
[19] "Vd2+"              "Vd2-"              "Va7.2+"           
[22] "Va7.2-"            "CD4+"              "CD4-"             
[25] "CD8+"              "CD8-"              "Tcells_count"     
[28] "lymphocytes_count" "Monocytes"         "Debris"           
[31] "CD45_count"
#it seems like metadata, flowcytometry names, and gates
data_alternative <- read.csv(file=thefilepath, check.names=TRUE)
colnames(data_alternative)
# the difference between check.names False or True. in False it 
#will give you names as they are
# "Va7.2..1" instead Var7.2-. 
# in True it will change special characters, spaces with a dot and + or - for 1 and 2.
# it convert in R names. R-approved syntax.
#you can convert the names to avoid special characters
#so for csv files always do FALSE in check.names
head # to see first rows
head(data, 5) # first name of the file and then number of rows you want to see.
View # to see hole matrix in "exel format"
#buy expanding in the second sidebar
class # to see which type is the whole thing
> class(data)
[1] "data.frame"
str # to see which type is each columm (internal structure)
#data.frame':	196 obs. of  31 variables
#also another way to check dataframes is the glimpse function
#but first first check in which package is this function
??glimpse # andthe help function tell that is dplyr
library(dplyr)
glimpse # very similar to str
glimpse(data)
# for each row it tells you which kind data is it. chr, int, dbl
#dbl = has decimal point
#int = integrity number/whole
#chr = letters.
dim # to see number of rows + number of colums
 dim(data)
[1] [196  31] # first rows,196 and then colums,31
#to find somthing inside a the data set for colums
data$Tcells_count. # the file, and name of the columm
#to check which kind of data is it
is.character(data$Tcells_count) #False
is.numeric(data$Tcells_count) # True
is.integer(data$Tcells_count) # for hole numbers
is.double(data$Tcells_count) # for numbers with decimals.
#special characters, need tick marks
is.numeric(data$CD8-) #instead
is.numeric(data$`CD8-`)
library(dplyr) 
# used to manipulate and clean tabular data
#select colums, filter rows, create new variables, group data, summarise data
#select, one colum from all the data set. 

datecolumm <-select (data, "Date") # dataset, and desired columm
# or you can write it like this too
datecolumm1 <- data  |> select("Date")
#this is a pipe   %>% 
#which is the same as this one next
datecolumm2 <- data  %>% select("Date")
#pipes allow to link togheter differnt ones instead of needeing
#to write line by line
#can be linked together, passing resulting objects of one operation on to the next pipe and subsequent function.

nrow() # total number of rows
data |> select("Date") |> nrow()
# so in this case will be how many rows  there is in the colum "date"
data |> select("Date") |> head(5)
#will give you the first 5 rows of  the "date" data
#select multiples columms. 
subtset <-data |> select(Tcells, `CD8+`, `CD4+`)
subtset2 <-data |> select (`CD8+`, `CD4+`, bid)
#you can select the colums that you want
relocate() # to change the colums around
# does not chage the csv file, just how we see in R session.

data |> relocate(Tcells, .after=bid)|>head(5)
data |> relocate(Tcells, Monocytes, .after=bid)|>head(5)
data |> relocate(Tcells, .before=Date)|>head(5)
# you can put befre or after a colum name but also considering
#the postion like 1, 3...
data |> relocate(Tcells, .before=2)|>head(5)
# also to just select one colum
select()
data |> select(singletsSSCB) # will show just information for this colum
#to rename
colnames(subset)
colnames(subset)[3] <- CD4positive
#alternative you can use renamed
ranemed <-subset |> rename (CD8pos =`CD8+`)
#first the new name and then old name
pull # grave the numbers from a specific columm
#ex: just the dates for the 4 first rows
data |> pull(Date)|> head(4)
data |> pull(lymphocytes) |> head (5)
data |> pull(`Va7.2-`) |> head (5)
unique # will give you the non repeated values
data |> pull(Date)|> unique()
#now for rows
 data |> pull(Condition) |>unique()
[1] "Ctrl" "PPD"  "SEB"
#filter
PPDonly <- data |> filter(Condition == "PPD")
# == matching this name
# in == sign can have glitshes insted %in%
# %in% is safer approach when you want to use character values/numerics
data |> filter(Condition %in% "PPD") |> head(10)
#for specify multiples you create a vector as an object
data |> filter(Condition %in% c("PPD","SEB")) |> head(10)
# you can do an object for that
theseconditions <- c("PPD","SEB")
data |> filter(Condition %in% theseconditions) |> head(10)
# for Conditionals
# using the exclamation mark will change the value to opossite
# isthisspectral <-TRUE
# !isthisspectral = FALSE
IsThisASpectralInstrument <- TRUE
!IsThisASpectralInstrument
[1] FALSE
subset <- data |> filter(!Conditions %in% "SEB")
# this will keep everything less SEB
#timepoints 0,4,9
subset_2 <- data |> filter (timepoint%in% "0" )# just for 0h
subset_2 <- data |> filter (!timepoint%in% "0" ) # for 4 and 9h only, not 0
#mutate
TidyData <- Data |> filter(Condition %in% "Ctrl") |> filter(timepoint %in% "0") |>
     select(bid, timepoint, Condition, Date, Tcells_count, CD45_count) |>
      rename(specimen=bid, condition=Condition) |> relocate(Date, .after=specimen)
    #first condition colum and just selecting control, then the time point
    #where just value of 0h
    # then select the number of colums
    #then rename to change the colum names
    #relocate, place date after specifmen colum
#so now output is just 23 rows
    specimen       Date timepoint condition Tcells_count CD45_count
1   INF0052 26/07/2025         0      Ctrl       164771     915203
2   INF0100 26/07/2025         0      Ctrl       208241    1438047
3   INF0179 26/07/2025         0      Ctrl       291777     940733
4   INF0134 29/07/2025         0      Ctrl       127866     689676
5   INF0148 29/07/2025         0      Ctrl       234335    1013985
6   INF0191 29/07/2025         0      Ctrl        55780     715443
7   INF0124 31/07/2025         0      Ctrl        70297     687720
8   INF0149 31/07/2025         0      Ctrl       107900     857845
9   INF0169 31/07/2025         0      Ctrl        75540     854594
10  INF0019 05/08/2025         0      Ctrl       208055     873622
11  INF0032 05/08/2025         0      Ctrl       361034     753064
12  INF0180 05/08/2025         0      Ctrl       284958    1049663
13  INF0155 07/08/2025         0      Ctrl       281626    1065048
14  INF0158 07/08/2025         0      Ctrl       280913    1249338
15  INF0159 07/08/2025         0      Ctrl       452551    1190219
16  INF0013 22/08/2025         0      Ctrl       182751     836573
17  INF0023 22/08/2025         0      Ctrl       218435     968035
18  INF0030 22/08/2025         0      Ctrl        85521     732321
19  INF0166 28/08/2025         0      Ctrl       225650     739495
20  INF0199 28/08/2025         0      Ctrl       169736    1112176
21  INF0207 28/08/2025         0      Ctrl        39055     905365
22  INF0614 30/08/2025         0      Ctrl       224396    1569007
23  INF0622 30/08/2025         0      Ctrl       161924     939307
#mutate() function can be used to modify existing columns, 
# as well as to create new ones
# so to get the proportion of T cells
TidyData <- TidyData |> mutate(Tcells_ProportionCD45 = Tcells_count / CD45_count)
# so new colum will have a algorithm in this case just 
#a division so colum = algorith
# but it give too many decimal digits 0.18003765
# so to ruond them up
TidyData <- TidyData |> mutate(TcellsRounded = round(Tcells_ProportionCD45, 2))
# so new colum will apear
#or if we want to do percentatge
TidyData <- TidyData |> mutate(TcellsPercentage = TcellsRounded * 100)
#arrange
#this is for rows
desc() or arrange()
TidyData <- TidyData |> arrange(desc(TcellsRounded))
TidyData |> filter(TcellsRounded > 0.3)
TidyData |> filter(TcellsRounded > 0.3) |> pull(specimen)
[1] "INF0032" "INF0159" "INF0179" "INF0166"
#so convination of each function to see data you need
#you may want to save this for later use
#istead of modifying the existing csv file you can create a new one
NewName <- paste0("MyNewDataset", ".csv")
StorageLocation <- file.path("data", NewName)
write.csv(TidyData, StorageLocation, row.names=FALSE)

#this way is easier to track all changes that you are doing in the file
# you don't overight data


git add .
git commit -m "end HW"
git push