Part I: Data inspection

Loading (tidyverse) package:

library(tidyverse)

Downloading files

downloaded files from github into Assignment2 folder

genotypes_data<-read_tsv("fang_et_al_genotypes.txt")

snp_data<-read_tsv("snp_position.txt")

Data Inspection

Load data:

str(genotypes_data) str(snp_data) unique(genotypes_data$Group) unique(snp_data$Chromosome) nrow(genotypes_data) nrow(snp_data) ncol(dgenotypes_data) ncol(snp_data) dim(genotypes_data) dim(snp_data)

Data Processing

#Rearranged the snp_data file so that the Chromosome column is followed by position, followed by transposition

snp_data<-snp_data[c(1,3,4,2,5:15)]

#Subset the genotypes_data dataframe for the maize and teosinte group seperately

maize_genotypes_data<-genotypes_data %>% filter(Group=="ZMMIL"|Group=="ZMMLR"|Group=="ZMMMR")

teosinte_genotypes_data<-genotypes_data %>%filter(Group=="ZMPJA"|Group=="ZMPIL"|Group=="ZMPBA")

Transposed the subset of genotypes_data to merge with the snp_data data frame

maize_genotypes_data<-as.data.frame(t(maize_genotypes_data), stringsAsFactors = F) teosinte_genotypes_data<-as.data.frame(t(teosinte_genotypes_data), stringsAsFactors = F)

#Column names transposed to row names. Created another row from the row names called SNP_ID. First row converted to row names and removed the first three columns. SNP_ID <- rownames(maize_genotypes_data) rownames(maize_genotypes_data) <- NULL maize_genotypes_data<- cbind(SNP_ID,maize_genotypes_data,stringsAsFactors = FALSE) names(maize_genotypes_data)<- c("SNP_ID",maize_genotypes_data[1,-1]) maize_genotypes_data <- maize_genotypes_data[-c(1,2,3),]

SNP_ID <- rownames(teosinte_genotypes_data) rownames(teosinte_genotypes_data) <- NULL teosinte_genotypes_data<- cbind(SNP_ID,teosinte_genotypes_data,stringsAsFactors = FALSE) names(teosinte_genotypes_data)<- c("SNP_ID",teosinte_genotypes_data[1,-1]) teosinte_genotypes_data <- teosinte_genotypes_data[-c(1,2,3),]

#merged the transposed genotype file with snp file

merged_maize<-merge(snp_data, maize_genotypes_data, by="SNP_ID") merged_teosinte<-merge(snp_data, teosinte_genotypes_data, by="SNP_ID") data<-merged_maize %>% filter(Chromosome %in% c(1))

#subset the merged file according to Chromosome number(numerical values) then sorted it based on position(ascending values). Exported as csv file based on Chrom number, order and grouping(maize/teosinte) for(i in c(1:10)){ data<-merged_maize %>% filter(Chromosome==i)%>% mutate(Pos=as.numeric(Position))%>%arrange(Pos) data$Pos<-NULL write.csv( data,paste0("Maize_Chromosome_",i,"ascending.csv"), row.names = F) data<-merged_maize%>% filter(Chromosome==i)%>% mutate(Pos=as.numeric(Position))%>% arrange(-Pos) data$Pos<-NULL data[data=="?/?"]<-"-/-" write.csv( data,paste0("Maize_Chromosome",i,"_descending.csv"), row.names = F)

data<-merged_teosinte %>% filter(Chromosome==i)%>% mutate(Pos=as.numeric(Position))%>%arrange(Pos) data$Pos<-NULL write.csv( data,paste0("Teosinte_Chromosome_",i,"ascending.csv"), row.names = F) data<-merged_teosinte %>% filter(Chromosome==i)%>% mutate(Pos=as.numeric(Position))%>% arrange(-Pos) data$Pos<-NULL data[data=="?/?"]<-"-/-" write.csv( data,paste0("Teosinte_Chromosome",i,"_descending.csv"), row.names = F)

}

##Part 2

#Data Visualization

#loaded ggplot library(ggplot2)

#visualized the number of polymorphism positions in each chromosome

ggplot(data = snp_data[!is.na(as.numeric(snp_data$Chromosome)),]) + geom_bar(mapping = aes(x = as.numeric(Chromosome), fill=Chromosome)) + scale_x_discrete(limit=c(1:10))+ labs(x = "Chromosome", y="No. of polymorphism position")

#tidying up the data using pivot_longer

#loaded tidyr

library(tidyr)

tidy_file<-genotypes_data %>% pivot_longer(!c("Sample_ID", "JG_OTU", "Group"),names_to = "SNP_ID",values_to = "base")

tidy_file<-merge(tidy_file, snp_data, by="SNP_ID")

ggplot(data = tidy_file[!is.na(as.numeric(tidy_file$Chromosome)),]) + geom_bar(mapping = aes(x = as.numeric(Chromosome), fill=Group)) + scale_x_discrete(limit=c(1:10))+ labs(x = "Chromosome" , y="No. of polymorphism position")

#Representation of polymorphisms contributed by each group

graph <- tidy_file

#Mutate Chromosone number into numeric mutate(Chromosome=as.numeric(Chromosome)) %>%

#Select required fields select(Group, SNP_ID, Chromosome, base) %>%

#Filter chromosome 1 to 10 filter(Chromosome %in% c(1:10)) %>%

#remove all unknown bases
filter(base!="?/?")%>%

#group by Group and SNP_ID
group_by(Group,SNP_ID) %>%

#remove all SNPS in a group that has only one base i.e no polymorphism in the group filter(length(unique(base))>1) %>%

#select requred fields select(Group, SNP_ID, Chromosome)

graph<-graph[!duplicated(graph),] #remove duplicated records

#plotted the filtered data based on polymorphism contributed by each group ggplot(data=graph) + geom_bar(mapping=aes(x=Chromosome, fill=Group)) + scale_x_discrete(limit=c(1:10), label=c(1:10))

#created a new column based on homozygousity of the polymorphism

tidy_file$homozygous<-TRUE tidy_file$homozygous[tidy_file$base=="?/?"]<-NA

tidy_file$homozygous[substr(tidy_file$Base,1,1)!=substr(tidy_file$base,3,3)]<-FALSE

Graphed the SNPs by Group, filling by Homozygosity:

graphs<-tidy_file ggplot(data=tidy_file) + geom_bar(mapping=aes(x=Group, fill=homozygous), position="fill")

#Graphs based on polymorphism in every chromosome fig.width=10 fig.height=11

SNP_graph<-tidy_file[!is.na(as.numeric(tidy_file$Chromosome)),] SNP_graph$Chromosome<-as.numeric(SNP_graph$Chromosome) SNP_graph<-SNP_graph%>% select(SNP_ID, Group, base, Position, Chromosome, homozygous)%>% filter(base!="?/?")%>% unique()%>%filter(!is.na(as.numeric(Position)))

final_graph<-ggplot(data = SNP_graph) + geom_point(mapping=aes(x=as.numeric(Position), y=Group, color=homozygous)) +labs(y = "Group" , x="Chromosome position")

final_graph + facet_wrap(~ Chromosome,ncol=2, scales = "free") +labs(color="Genotype")