submit_Xiyuan_Guan

analysis.R

@@ -0,0 +1,101 @@
+# Introduction to Biocomputing - R Project
+# Written by Xiyuan Guan
+# 2022-12-1
+# analysis
+library(ggplot2)
+
+setwd("C:/Users/guanx/Documents/N.ND/BIOS 60318 Biocomputing/test/Rproject/Rproject/")
+source("supportingFunctions.R")
+
+# converts all .txt files in "countryY" into .csv format.####
+trans.csv("countryY"," ")
+
+
+# compile data from all .csv files in "countryX" and "countryY"
+sumtable1 <- combine.csv("countryX","Warn")
+sumtable2 <- combine.csv("countryY","Warn")
+# compile all data
+allcsv <- rbind(sumtable1,sumtable2)
+write.csv(allcsv,"allData_myVersion.csv", row.names = F)
+
+# summarize the compiled data set
+file <- read.csv("allData.csv")
+sub.pat <- sumdata(file) # this step takes time
+# the summary of the age distribution of patients is shown in console. 
+write.csv(sub.pat,"subset_of_patients.csv",row.names = F) # for better load
+
+# beginning of answers####
+
+# Q1: In which country (X or Y) did the disease outbreak likely begin?####
+# generate dataset with day, number of patients and country
+day_num <- data.frame(matrix(NA,0,3))
+colnames(day_num) <- c("dayofYear","num_patient","country")
+Xdays<- unique(file$dayofYear[which(file$country=="X")]) # days with patients in X country
+Ydays <- unique(file$dayofYear[which(file$country=="Y")])
+setdiff(Xdays,Ydays) # same
+for(i in 1:length(Xdays)){
+  day_num[i,1] <- Xdays[i] # dayofYear
+  XPat <- file[which(file$dayofYear==Xdays[i]&file$country=="X"),]# subset country and day
+  nXPat<- XPat[which(rowSums(XPat[,3:12])>0),] # subset patients
+  day_num[i,2] <- nrow(nXPat) # num of patient
+  day_num[i,3] <- "X"
+}
+nrow(day_num) #56
+for(i in 1:length(Ydays)){
+  day_num[i+56,1] <- Ydays[i] # dayofYear
+  YPat <- file[which(file$dayofYear==Ydays[i]&file$country=="Y"),] # subset country and day
+  nYPat<- YPat[which(rowSums(YPat[,3:12])>0),] # subset patients
+  day_num[i+56,2] <- nrow(nYPat) # num of patient
+  day_num[i+56,3] <- "Y"
+}
+
+head(day_num)
+
+# sum(day_num[,2]) # test data accuracy = nrow(sub.pat)
+
+# save as plot_1.png
+ggplot(data=day_num, aes(x = dayofYear, y = num_patient, group = country, color = country, shape = country))+
+  geom_line(linewidth = 0.1)+
+  geom_point()+
+  theme_bw()
+
+# accumulated num of patient could be better
+# generate dataset with day, accumulated number of patients, country
+day_num$acc_num[1:56] <- cumsum(day_num$num_patient[1:56]) # 
+day_num$acc_num[57:112] <- cumsum(day_num$num_patient[57:112]) #Y
+
+# save as plot_2.png
+ggplot(data=day_num, aes(x = dayofYear, y = acc_num, group = country, color = country, shape = country))+
+  geom_line(linewidth = 0.5)+
+  geom_point(size=1)+
+  theme_bw()
+
+# A1: According to plot_1 and plot_2, the disease started to break in country X before the 120th day of the year,
+# which is the start day of screening. However, the disease started to break in country y on around the 139th day
+# of the year, so the disease outbreak likely begin in country X.
+
+# Q2: If Country Y develops a vaccine for the disease, is it likely to work for citizens of Country X?####
+
+mark_num <- data.frame(matrix(NA,20,3))
+colnames(mark_num) <- c("marker","num_of_patient","country")
+mark_num[,1] <- rep(colnames(file[,3:12]),2) # 10 markers
+mark_num[1:10,3] <- "X"
+mark_num[11:20,3] <- "Y"
+for(i in 3:12){
+  mark_num[i-2,2] <- sum(file[which(file$country=="X"),i])
+  mark_num[i+8,2] <- sum(file[which(file$country=="Y"),i])
+}
+
+#generate plot_3
+ggplot(data = mark_num, aes(x = marker, y = num_of_patient, fill = country))+
+  geom_bar(stat = "identity", position = "dodge")+
+  theme(axis.text.x = element_text(angle = 60, hjust = 1))+
+  scale_y_continuous(expand = c(0,0))+
+  xlab("")
+
+# A2: According to plot_3, in country X, marker01-05 are mainly present in patients' sample,
+# while in country Y, marker04-10 are mainly present, which indicates different proteins
+# shown to form the immunological response by the patients, so indicates differences in the 
+# protein in the disease. So the vaccine developed by Country Y may work on the strain that
+# has marker04-10 as the response to disease-causing agents, and may not work on the strain 
+# in Country X with marker01-05 as the response.