PBGVPedigree Cleaning.Rmd

---
title: "PBGV Pedigree Cleaning"
output: html_document
date: '2023-04-12'
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)

# Clean session
rm(list = ls())

# Install packages
library(tidyverse)
library(skimr)
library(optiSel)

# Install AlleleFetch - for function SortPedigree()
if (FALSE) {
  setRepositories() # choose CRAN and BioC
  install.packages(pkg = c("gRain", "devtools"), dep = TRUE)
  # PR that fixed the repeatPattern() data issue
  # https://github.com/hojsgaard/gRain/pull/9
  devtools::install_github(repo = "https://github.com/gregorgorjanc/gRain")
}
# Load package
library(package = "gRain")

# User defined functions

source(file = "AlleleFetch.R")

```

## Import the raw dataset
```{r}
setwd(dir="/Users/roscraddock/Documents/University\ of\ Edinburgh/HighlanderLab - AlleleFetch_data")

PBGV_POAG <- read.csv(file = "PBGV_POAG_data.csv")
```

## Correcting the Data Formats
Check the formatting of the data
```{r}
str(PBGV_POAG)
```

Correct the formatting
```{r}
PBGV_POAG$IRN <- as.character(PBGV_POAG$IRN)
PBGV_POAG$IRN_sire <- as.character(PBGV_POAG$IRN_sire)
PBGV_POAG$IRN_dam <- as.character(PBGV_POAG$IRN_dam)
PBGV_POAG$Sex <- as.factor(PBGV_POAG$Sex)
PBGV_POAG$DOB <- as.Date(PBGV_POAG$DOB, "%Y-%m-%d")

```

## Pedigree Cleaning

Check the levels of sex - either dog or bitch
```{r}
unique(PBGV_POAG$Sex)
```
-> Two levels as expected, no correction needed.


Check if all recorded sires and dams have their own records in the pedigree
```{r}
test <- PBGV_POAG %>%
  filter(!is.na(IRN_sire) & !IRN_sire %in% IRN)
count(test)

test <- PBGV_POAG %>%
  filter(!is.na(IRN_dam) & !IRN_dam %in% IRN)
count(test)
```
-> All recorded sires and dams have their own records, no correction needed.


Check if all sires are dogs, and all dams are bitches.
```{r}
test <- PBGV_POAG %>%
  filter(!is.na(IRN) & IRN %in% IRN_sire)

#Identify any bitches
print(test[test$Sex=="Bitch",])
#2 sires recorded are said to be female.

test <- PBGV_POAG %>%
  filter(!is.na(IRN) & IRN %in% IRN_dam)

#Identify any dog
print(test[test$Sex=="Dog",])
#2 dams recorded are said to be male.
```

To correct, the recorded sire or dam are removed where the sex does not match
```{r}
#6790410 - recorded as female, but said to be sire for two dogs. Removed from parents
PBGV_POAG <- PBGV_POAG %>%
  transform(IRN_sire=ifelse(IRN_sire=="6790410", NA, IRN_sire))

#453000
PBGV_POAG <- PBGV_POAG %>%
  transform(IRN_sire=ifelse(IRN_sire=="453000", NA, IRN_sire))

#453183
PBGV_POAG <- PBGV_POAG %>%
  transform(IRN_dam=ifelse(IRN_dam=="453183", NA, IRN_dam))

#453207
PBGV_POAG <- PBGV_POAG %>%
  transform(IRN_dam=ifelse(IRN_dam=="453207", NA, IRN_dam))

```

Check all has worked as expected
```{r}
test<- PBGV_POAG %>%
  filter(!is.na(IRN) & IRN %in% IRN_sire)

#Identify any bitches
print(test[test$Sex=="Bitch",])


test <- PBGV_POAG %>%
  filter(!is.na(IRN) & IRN %in% IRN_dam)

#Identify any dog
print(test[test$Sex=="Dog",])

```


Test for any duplicates - all due to multiple eye tests
```{r}

sum(duplicated(PBGV_POAG$IRN))
#63 duplicates found

test <- PBGV_POAG[duplicated(PBGV_POAG$IRN),]

#6841428 - remove duplicate to display most recent eye test
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "6841428" & PBGV_POAG$EyeTest_Result== "Unaffected"),]

#6841595
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "6841595" & PBGV_POAG$EyeTest_Date== "2016-06-25"),]

#6862679 - previously diagnosed as affected, but later unaffected. Assumed due to successful treatment so selected affected to remove treatment. 
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "6862679" & !PBGV_POAG$EyeTest_Date== "2017-03-02"),]

#7233110
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "7233110" & PBGV_POAG$EyeTest_Date== "2016-04-07"),]

#7609612
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "7609612" & !PBGV_POAG$EyeTest_Date== "2018-02-22"),]

#7733685
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "7733685" & PBGV_POAG$EyeTest_Date== "2019-04-20"),]

#7733686 
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "7733686" & !PBGV_POAG$EyeTest_Date== "2021-02-23"),]

#7785875
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "7785875" & PBGV_POAG$EyeTest_Date== "2014-03-15"),]

#7852467
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "7852467" & PBGV_POAG$EyeTest_Date== "2014-03-15"),]

#8044309
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8044309" & PBGV_POAG$EyeTest_Result== "Test results with owner"),]

#8593652
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8593652" & !PBGV_POAG$EyeTest_Date== "2019-10-24"),]

#8616739
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8616739" & !PBGV_POAG$EyeTest_Date== "2020-02-06"),]

#8616742
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8616742" & PBGV_POAG$EyeTest_Date== "2014-11-28"),]

#8685310
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8685310" & !PBGV_POAG$EyeTest_Date== "2022-10-25"),]

#8822546
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8822546" & !PBGV_POAG$EyeTest_Date== "2019-04-20"),]

#8827943
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "827943" & PBGV_POAG$EyeTest_Date== "2014-08-21"),]

#8888576 - latest says unaffected, again assumed from successful treatment as previously identified affected. 
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "888576" & PBGV_POAG$EyeTest_Date== "2020-12-01"),]

#8980798
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8980798" & PBGV_POAG$EyeTest_Date== "2016-08-03"),]

#8980799
PBGV_POAG <-PBGV_POAG[!(PBGV_POAG$IRN == "8980799" & !PBGV_POAG$EyeTest_Date== "2019-11-15"),]

#8951848
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8951848" & PBGV_POAG$EyeTest_Date== "2014-11-27"),]

#9144292
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "9144292" & PBGV_POAG$EyeTest_Date== "2020-09-08"),]

#9575670 - most recent eye test no result recorded, so using the second most recent.
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "9575670" & PBGV_POAG$EyeTest_Date== "2022-11-11"),]

#9668778
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "668778" & PBGV_POAG$EyeTest_Date== "2020-03-06"),]

#9878058
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "9878058" & !PBGV_POAG$EyeTest_Date== "2022-11-11"),]

#9878920
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "9878920" & PBGV_POAG$EyeTest_Date== "2018-10-20"),]

#9943273
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "9943273" & PBGV_POAG$EyeTest_Date== "2020-09-08"),]

#10123773
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "10123773" & PBGV_POAG$EyeTest_Date== "2019-04-20"),]

#8202645
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8202645" & !PBGV_POAG$EyeTest_Date== "2020-02-06"),]

#8827943
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8827943" & PBGV_POAG$EyeTest_Date== "2014-08-21"),]

#8888576 - latest says unaffected, again assumed from successful treatment as previously identified affected.
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "8888576" & PBGV_POAG$EyeTest_Date== "2020-12-01"),]

#9668778
PBGV_POAG <- PBGV_POAG[!(PBGV_POAG$IRN == "9668778" & PBGV_POAG$EyeTest_Date== "2020-03-06"),]


```

Check that all duplicates have been corrected
```{r}
sum(duplicated(PBGV_POAG$IRN))
```


Remove individuals with no sire, dam, or offspring recorded - they introduce error, and have no genotype or phenotype recorded.
```{r}
For_Cleaning <- PBGV_POAG %>%
  select(IRN, IRN_sire, IRN_dam)

head(For_Cleaning)
Pedig <- prePed(For_Cleaning)

#Identified the 42 observations that have no sire, dam, and no offspring
tmp <- Pedig %>%
  filter(is.na(Sire)) %>%
  filter (is.na(Dam)) %>%
  filter(Offspring== "FALSE")


#Remove from PBGV_POAG for analysis
tmp2 <- PBGV_POAG$IRN %in% tmp$Indiv
PBGV_POAG <- PBGV_POAG[!tmp2,]

```

Save the cleaned pedigree and dataset
```{r}
write.csv(PBGV_POAG, "/Users/roscraddock/Documents/University of Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/FullCleanedData.csv")
```


Create separate dataset that does not have hereditary status
```{r}
PBGV_POAG_DNA <- PBGV_POAG %>%
  transform(DNA_result=ifelse(DNA_result=="Hereditary Clear", NA, DNA_result))

PBGV_POAG_DNA <- PBGV_POAG_DNA %>%
  transform(DNA_result=ifelse(DNA_result=="Hereditary Carrier", NA, DNA_result))

PBGV_POAG_DNA <- PBGV_POAG_DNA %>%
  transform(DNA_result=ifelse(DNA_result=="Hereditary Affected", NA, DNA_result))
```


## Prepare files for full pedigree imputation in AlphaPeel

Need two files: 
     Genotype file (list of ID with genotypes either 0, 1, 2, 9 (for unknown)) and 
     Pedigree file, 0 where individual is unknown.
```{r}
#Separate ID, FID, MID, Genotype
Alpha_Peel <- PBGV_POAG_DNA %>%
  select(IRN, IRN_sire, IRN_dam, DNA_result, DOB, DNA_TestDate)

Alpha_Peel <- Alpha_Peel %>%
  rename(IID = "IRN", FID = "IRN_sire", MID="IRN_dam", Genotype = "DNA_result")

#Reorder pedigree so founders, then offspring
Alpha_Peel <- SortPedigree (Alpha_Peel)

#Recode pedigree
Alpha_Peel_Recode <- data.frame(IID = 1:nrow(Alpha_Peel),
             FID = match(x = Alpha_Peel[[2]], table = Alpha_Peel[[1]], nomatch = NA),
             MID = match(x = Alpha_Peel[[3]], table = Alpha_Peel[[1]], nomatch = NA))

#Combine the genotype, DOB, and DNA_TestDate with the recoded pedigree
tmp <- Alpha_Peel %>%
  select(DOB, Genotype, DNA_TestDate)
Alpha_Peel_Recode <- cbind(Alpha_Peel_Recode, tmp)

#Save for later plots and analysis
write.csv(Alpha_Peel_Recode, "/Users/roscraddock/Documents/University\ of\ Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/fullPBGV_Alpha_Peel_Recode.csv")

#Separate Pedigree
Alpha_Peel_Ped <- Alpha_Peel_Recode %>%
  select(IID, FID, MID)
#Change any NA to 0 (in FID and MID)
Alpha_Peel_Ped <- Alpha_Peel_Ped %>%
  transform(FID=ifelse(is.na(FID), 0, FID))
Alpha_Peel_Ped <- Alpha_Peel_Ped %>%
  transform(MID=ifelse(is.na(MID), 0, MID))

#Separate into Genotype
Alpha_Peel_Geno <- Alpha_Peel_Recode %>%
  select(IID, Genotype)
#Change any NA to 9 
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(Genotype=="", 9, Genotype))
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(is.na(Genotype), 9, Genotype))
#Change Clear to 0, Carrier to 1, Affected to 2
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(Genotype=="Clear", 0, Genotype))
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(Genotype=="Carrier", 1, Genotype))
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(Genotype=="Affected", 2, Genotype))
Alpha_Peel_Geno$Genotype <- as.numeric(Alpha_Peel_Geno$Genotype)

```


Save into text formats
```{r}
write.table(Alpha_Peel_Ped, "/Users/roscraddock/Documents/University\ of\ Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/AlphaPeel Inputs/PBVG_Ped.txt", row.names = FALSE, col.names = FALSE)

write.table(Alpha_Peel_Geno, "/Users/roscraddock/Documents/University\ of\ Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/AlphaPeel Inputs/PBVG_Geno.txt", row.names = FALSE, col.names = FALSE)
```

Separating the hereditary status for validation of AlphaPeel
```{r}
#Separate ID, FID, MID, Genotype
Alpha_Peel <- PBGV_POAG %>%
  select(IRN, IRN_sire, IRN_dam, DNA_result, DOB)

Alpha_Peel <- Alpha_Peel %>%
  rename(IID = "IRN", FID = "IRN_sire", MID="IRN_dam", Genotype = "DNA_result")

#Reorder pedigree so founders, then offspring
Alpha_Peel <- SortPedigree (Alpha_Peel)

#Recode pedigree
Alpha_Peel_Recode <- data.frame(IID = 1:nrow(Alpha_Peel),
             FID = match(x = Alpha_Peel[[2]], table = Alpha_Peel[[1]], nomatch = NA),
             MID = match(x = Alpha_Peel[[3]], table = Alpha_Peel[[1]], nomatch = NA))

#Insert the DNA results back in
Alpha_Peel <- Alpha_Peel %>%
  select(Genotype, DOB)
Val_Alpha <- cbind(Alpha_Peel_Recode, Alpha_Peel)

write.csv(Val_Alpha, "/Users/roscraddock/Documents/PBGV_Full_Val_Alpha.csv")
```

## Subsetting the pedigree
Scale testing was completed, separating the pedigree by dates of birth.
The optimum to test the hypothesis was found between January 2009 and June 2014.
```{r}
# Separating from 2009 to June 2014 by date of birth
PBGV_POAG_DNA$DOB <- as.Date(PBGV_POAG_DNA$DOB)
PBGV_POAG_Sub <- PBGV_POAG_DNA[!PBGV_POAG_DNA$DOB < "2008-12-31",]
PBGV_POAG_Sub <- PBGV_POAG_Sub[!PBGV_POAG_Sub$DOB > "2014-06-30",]
# Adding individuals with no date of birth
tmp <- PBGV_POAG_DNA[is.na(PBGV_POAG_DNA$DOB),]
PBGV_POAG_Sub <- rbind(PBGV_POAG_Sub, tmp)
#Removing any observations with an IRN of NA
PBGV_POAG_Sub <- PBGV_POAG_Sub[!is.na(PBGV_POAG_Sub$IRN),]

#If any sire without own record in pedigree, set to NA
test <- PBGV_POAG_Sub %>%
  filter(!is.na(IRN_sire) & !IRN_sire %in% IRN)
test$IRN_sire <- NA

tmp <- PBGV_POAG_Sub %>%
  filter(!is.na(IRN_sire) & IRN_sire %in% IRN)
tmp2 <- PBGV_POAG_Sub %>%
  filter(is.na(IRN_sire))
PBGV_POAG_Sub <- rbind(tmp, tmp2, test)

#If any dam without own record in pedigree, set to NA
test <- PBGV_POAG_Sub %>%
  filter(!is.na(IRN_dam) & !IRN_dam %in% IRN)
test$IRN_dam <- NA

tmp <- PBGV_POAG_Sub %>%
  filter(!is.na(IRN_dam) & IRN_dam %in% IRN)
tmp2 <- PBGV_POAG_Sub %>%
  filter(is.na(IRN_dam))
PBGV_POAG_Sub <- rbind(tmp, tmp2, test)

#Remove any observations with IRN of NA
PBGV_POAG_Sub <- PBGV_POAG_Sub[!is.na(PBGV_POAG_Sub$IRN),]

#Check the pedigree with OptiSel
For_Cleaning <-PBGV_POAG_Sub %>%
  select(IRN, IRN_sire, IRN_dam)

head(For_Cleaning)
Pedig <- prePed(For_Cleaning)

#Select individuals with no recorded dam, sire, or offspring
tmp <- Pedig %>%
  filter(is.na(Sire)) %>%
  filter (is.na(Dam)) %>%
  filter(Offspring== "FALSE")

#Remove from PBGV_POAG_Sub as introduces error
tmp2 <- PBGV_POAG_Sub$IRN %in% tmp$Indiv
PBGV_POAG_Sub <- PBGV_POAG_Sub[!tmp2,]
```

Save to files
```{r}
write.csv(PBGV_POAG_Sub, "/Users/roscraddock/Documents/University\ of\ Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/PBGV_Sub_2009-2014.csv")
```

## Preparing the files for AlphaPeel
Separate into AlphaPeel Format
```{r}
# Separate ID, FID, MID, Genotype
Alpha_Peel <- PBGV_POAG_Sub %>%
  select(IRN, IRN_sire, IRN_dam, DNA_result, DOB, DNA_TestDate)

Alpha_Peel <- Alpha_Peel %>%
  rename(IID = "IRN", FID = "IRN_sire", MID="IRN_dam", Genotype = "DNA_result")

# Reorder pedigree so founders, then offspring
Alpha_Peel <- SortPedigree (Alpha_Peel)

# Recode pedigree
Alpha_Peel_Recode <- data.frame(IID = 1:nrow(Alpha_Peel),
             FID = match(x = Alpha_Peel[[2]], table = Alpha_Peel[[1]], nomatch = NA),
             MID = match(x = Alpha_Peel[[3]], table = Alpha_Peel[[1]], nomatch = NA))

# Re-combine the recoded pedigree with the genotype
tmp <- Alpha_Peel %>%
  select(DOB, Genotype, DNA_TestDate)
Alpha_Peel_Recode <- cbind(Alpha_Peel_Recode, tmp)

# Separate Pedigree
Alpha_Peel_Ped <- Alpha_Peel_Recode %>%
  select(IID, FID, MID)
#Change any NA to 0 (in FID and MID)
Alpha_Peel_Ped <- Alpha_Peel_Ped %>%
  transform(FID=ifelse(is.na(FID), 0, FID))
Alpha_Peel_Ped <- Alpha_Peel_Ped %>%
  transform(MID=ifelse(is.na(MID), 0, MID))

#Separate into Genotype
Alpha_Peel_Geno <- Alpha_Peel_Recode %>%
  select(IID, Genotype)
#Change any NA to 9 
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(Genotype=="", 9, Genotype))
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(is.na(Genotype), 9, Genotype))
#Change Clear to 0, Carrier to 1, Affected to 2
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(Genotype=="Clear", 0, Genotype))
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(Genotype=="Carrier", 1, Genotype))
Alpha_Peel_Geno <- Alpha_Peel_Geno %>%
  transform(Genotype=ifelse(Genotype=="Affected", 2, Genotype))
Alpha_Peel_Geno$Genotype <- as.numeric(Alpha_Peel_Geno$Genotype)
```

Save in text formats.
```{r}
#Save into text formats
write.table(Alpha_Peel_Ped, "/Users/roscraddock/Documents/University of Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/AlphaPeel Inputs/PBVG_Ped_2009-2014.txt", row.names = FALSE, col.names = FALSE)

write.table(Alpha_Peel_Geno, "/Users/roscraddock/Documents/University of Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/AlphaPeel Inputs/PBVG_Geno_2009-2014.txt", row.names = FALSE, col.names = FALSE)

#Save the recoded AlphaPeel Subset into an csv
write.csv(Alpha_Peel_Recode, "/Users/roscraddock/Documents/University\ of\ Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/Alpha_Peel_2009-2014.csv")
```


Prepare AlphaPeel Genotype files for LOOCV
```{r}
i <- 1
nid <- as.numeric(count(Alpha_Peel_Geno))
for (i in i:nid){
  tmp <- Alpha_Peel_Geno
  x <- Alpha_Peel_Geno[Alpha_Peel_Geno$IID == i,]
  test <- !x$Genotype == 9
  if (any(test)){
    tmp$Genotype[i] <- 9
    Filename <- paste("/Users/roscraddock/Documents/University of Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/AlphaPeel Inputs/", i, "PBVG_Geno_2009-2014.txt", sep="")
    write.table(tmp, file = Filename, row.names = FALSE, col.names = FALSE)
  }
}
```


## Preparing the files for AlleleFetch
```{r}
PedScale <- PBGV_POAG_Sub %>%
  select(IRN, IRN_sire, IRN_dam) %>%
  rename(IID="IRN", FID="IRN_sire", MID="IRN_dam")

GenoScale <- PBGV_POAG_Sub %>%
  select(IRN, DNA_result) %>%
  rename(IID="IRN", Genotype="DNA_result")

# Select only the individuals with genotypes
tmp <- GenoScale %>%
  filter(Genotype == "Clear")
tmp2 <-GenoScale %>%
  filter(Genotype == "Carrier")
tmp3 <-GenoScale %>%
  filter(Genotype == "Affected")

GenoScale <- rbind(tmp, tmp2, tmp3)

#Change the genotypes to "AA", "AB", "BB" notation
GenoScale <- GenoScale %>%
  transform(Genotype=ifelse(Genotype=="Clear", "AA", Genotype))
GenoScale <- GenoScale %>%
  transform(Genotype=ifelse(Genotype=="Carrier", "AB", Genotype))
GenoScale<- GenoScale %>%
  transform(Genotype=ifelse(Genotype=="Affected", "BB", Genotype))

# Check all the genotyped individuals are in the pedigree
test <-  !GenoScale$IID %in% PedScale$IID

#Select phenotypes to add to analysis.
PhenoScale <- PBGV_POAG_Sub %>%
  select(IRN, EyeTest_Result) %>%
  rename(IID="IRN", Phenotype="EyeTest_Result")

#Select only the individuals with phenotypes
tmp <- PhenoScale %>%
  filter(Phenotype== "Affected")
tmp2 <- PhenoScale %>%
  filter (Phenotype == "Unaffected")

PhenoScale <- rbind(tmp,tmp2)

#Change categories to "OK" and "NOK" from "Unaffected" and "Affected"
PhenoScale <-  PhenoScale %>%
  transform(Phenotype=ifelse(Phenotype=="Unaffected", "OK", "NOK"))
```

Save these into three separate files
```{r}
write.csv(GenoScale, "/Users/roscraddock/Documents/University of Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/AlleleFetch Inputs/PBGV_Geno_2009-2014.csv")
write.csv(PhenoScale, "/Users/roscraddock/Documents/University of Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/AlleleFetch Inputs/PBGV_Pheno_2009-2014.csv")
write.csv(PedScale, "/Users/roscraddock/Documents/University of Edinburgh/HighlanderLab - AlleleFetch_data/Clean Pedigrees/PBGV/AlleleFetch Inputs/PBGV_Ped_2009-2014.csv")
```