library(tidyverse)
# library(tidylog)
library(knitr)10 R Character Exercise
10.1 Load Libraries
10.2 Useful RStudio cheatsheet
See the “String manipulation with stringr cheatsheet” at
10.3 Scenario 1
You are working with three different sets of collaborators: 1) the clinical group that did the field work and generated the anthropometric measurements; 2) the medical laboratory that measured blood pressure in a controlled environment; and 3) the molecular laboratory that generated the genotypes.
clin <- read.table(file = "data/clinical_data.txt", header=TRUE)
kable(clin)| ID | height |
|---|---|
| 1 | 152 |
| 104 | 172 |
| 2112 | 180 |
| 2543 | 163 |
lab <- read.table(file = "data/lab_data.txt", header = TRUE)
kable(lab)| ID | SBP |
|---|---|
| SG0001 | 120 |
| SG0104 | 111 |
| SG2112 | 125 |
| SG2543 | 119 |
geno <- read.table(file = "data/genotype_data.txt", header = TRUE)
kable(geno)| Sample | rs1212 |
|---|---|
| TaqMan-SG0001-190601 | G/C |
| TaqMan-SG0104-190602 | G/G |
| TaqMan-SG2112-190603 | C/C |
| TaqMan-Sg2543-190603 | C/G |
10.4 Discussion Questions
10.4.1 Question 1
The clinical group, which measured height, used integer IDs, but the medical group, which measured the blood pressure, decided to prefix the integer IDs with the string ‘SG’ (so as to distinguish them from other studies that were also using integer IDs). So ID ‘1’ was mapped to ID ‘SG0001’.
| ID | height |
|---|---|
| 1 | 152 |
| 104 | 172 |
| 2112 | 180 |
| 2543 | 163 |
Discuss how, using R commands, you would reformat the integer IDs to be in the format “SGXXXX”. Write down your ideas in the next section, and, if you have time, try them out within an R chunk.
Hint: Use the formatC function.
10.4.1.1 Interactive WebR chunk
You can interactively run R within this WebR chunk by clicking the Run code tab. Note that this is a limited version of R which runs within your web browser.
Note
This Run code WebR chunk needs to be run first, before the later ones, as it downloads and reads in the required data files. The WebR chunks should be run in order, as you encounter them, from beginning to end.
10.4.2 Answer 1
Expand to see solution
clin$SUBJECT_ID <- paste0("SG", formatC(clin$ID, width = 4, flag = "0000"))
kable(clin)| ID | height | SUBJECT_ID |
|---|---|---|
| 1 | 152 | SG0001 |
| 104 | 172 | SG0104 |
| 2112 | 180 | SG2112 |
| 2543 | 163 | SG2543 |
# Or here's an alternative using the 'sub' command:
sub("00","SG",formatC(clin$ID, flag="0000", width=6)) [1] "SG0001" "SG0104" "SG2112" "SG2543"# Or can be done using a `case_when`:
case_when(
clin$ID < 10 ~ paste0("SG000",clin$ID),
clin$ID < 100 ~ paste0("SG00",clin$ID),
clin$ID < 1000 ~ paste0("SG0",clin$ID),
clin$ID < 10000 ~ paste0("SG",clin$ID)
)[1] "SG0001" "SG0104" "SG2112" "SG2543"10.4.3 Question 2
Discuss how, using R commands, you would reformat the “SGXXXX” IDs to be integer IDs. Write down your ideas in the next section, and, if you have time, try them out within an R chunk.
| ID | SBP |
|---|---|
| SG0001 | 120 |
| SG0104 | 111 |
| SG2112 | 125 |
| SG2543 | 119 |
Hint: Use either the gsub command or the str_replace_all command from the stringr package.
Warning
To read in and load the data within the WebR environment, be sure to run all of the WebR chunks in order. For example, to usefully run R code in this WebR chunk here, you first need to run the WebR chunk above in Question 1.
10.4.4 Answer 2
Expand to see solution
lab$ID2 <- as.numeric(gsub("SG","",lab$ID))
kable(lab)| ID | SBP | ID2 |
|---|---|---|
| SG0001 | 120 | 1 |
| SG0104 | 111 | 104 |
| SG2112 | 125 | 2112 |
| SG2543 | 119 | 2543 |
lab$ID2 <- NA
lab$ID2 <- str_replace_all(lab$ID, pattern = "SG", replacement = "") %>% as.numeric()
kable(lab)| ID | SBP | ID2 |
|---|---|---|
| SG0001 | 120 | 1 |
| SG0104 | 111 | 104 |
| SG2112 | 125 | 2112 |
| SG2543 | 119 | 2543 |
10.4.5 Question 3
The genotype group used IDs in the style “TaqMan-SG0001-190601”, where the first string is “TaqMan” and the ending string is the date of the genotyping experiment.
Discuss how, using R commands, you would extract an “SGXXXX” style ID from the “TaqMan-SG0001-190601” style IDs. Write down your ideas in the next section, and, if you have time, try them out within an R chunk.
Note that one of the IDs has a lower case ‘g’ in it - how would you correct this, using R commands?
| Sample | rs1212 |
|---|---|
| TaqMan-SG0001-190601 | G/C |
| TaqMan-SG0104-190602 | G/G |
| TaqMan-SG2112-190603 | C/C |
| TaqMan-Sg2543-190603 | C/G |
Hint: Use either the str_split_fixed function from the stringr package or the separate function from the tidyr package.
10.4.6 Answer 3
Expand to see solution
a <- str_split_fixed(geno$Sample, pattern = "-",n=3)
a [,1] [,2] [,3]
[1,] "TaqMan" "SG0001" "190601"
[2,] "TaqMan" "SG0104" "190602"
[3,] "TaqMan" "SG2112" "190603"
[4,] "TaqMan" "Sg2543" "190603"geno$ID <- toupper(a[,2])
kable(geno)| Sample | rs1212 | ID |
|---|---|---|
| TaqMan-SG0001-190601 | G/C | SG0001 |
| TaqMan-SG0104-190602 | G/G | SG0104 |
| TaqMan-SG2112-190603 | C/C | SG2112 |
| TaqMan-Sg2543-190603 | C/G | SG2543 |
The separate function from the tidyr package is also useful:
geno %>%
separate(Sample, into=c("Tech","ID2","Suffix"), sep="-") %>%
mutate(ID2=toupper(ID2)) Tech ID2 Suffix rs1212 ID
1 TaqMan SG0001 190601 G/C SG0001
2 TaqMan SG0104 190602 G/G SG0104
3 TaqMan SG2112 190603 C/C SG2112
4 TaqMan SG2543 190603 C/G SG2543The separate function is being superseded in favor of separate_wider_delim and separate_wider_position. In this case, separate_wider_delim is applicable.
geno %>%
separate_wider_delim(cols=Sample, delim = "-", names=c("Tech","ID2","Suffix")) %>%
mutate(ID2=toupper(ID2))# A tibble: 4 × 5
Tech ID2 Suffix rs1212 ID
<chr> <chr> <chr> <chr> <chr>
1 TaqMan SG0001 190601 G/C SG0001
2 TaqMan SG0104 190602 G/G SG0104
3 TaqMan SG2112 190603 C/C SG2112
4 TaqMan SG2543 190603 C/G SG254310.5 Scenario 2
A replication sample has been measured, and that is using IDs in the style “RP5XXX”.
joint <- read.table(file = "data/joint_data.txt", header = TRUE)
kable(joint)| ID | SBP |
|---|---|
| SG0001 | 120 |
| SG0104 | 111 |
| SG2112 | 125 |
| SG2543 | 119 |
| RP5002 | 121 |
| RP5012 | 118 |
| RP5113 | 112 |
| RP5213 | 142 |
10.5.1 Question 4
Discuss how you would use R commands to split the ‘joint’ data frame into an ‘SG’ and ‘RP’ specific piece? Write down your ideas in the next section, and, if you have time, try them out within an R chunk.
| ID | SBP |
|---|---|
| SG0001 | 120 |
| SG0104 | 111 |
| SG2112 | 125 |
| SG2543 | 119 |
| RP5002 | 121 |
| RP5012 | 118 |
| RP5113 | 112 |
| RP5213 | 142 |
10.5.2 Answer 4
Expand to see solution
grep(pattern = "SG",joint$ID)[1] 1 2 3 4grep(pattern = "RP", joint$ID)[1] 5 6 7 8joint.SG <- joint[grep(pattern = "SG",joint$ID), ]
joint.RP <- joint[grep(pattern = "RP", joint$ID), ]
kable(joint.SG)| ID | SBP |
|---|---|
| SG0001 | 120 |
| SG0104 | 111 |
| SG2112 | 125 |
| SG2543 | 119 |
kable(joint.RP)| ID | SBP | |
|---|---|---|
| 5 | RP5002 | 121 |
| 6 | RP5012 | 118 |
| 7 | RP5113 | 112 |
| 8 | RP5213 | 142 |
# Reset row names
rownames(joint.RP) <- NULL
kable(joint.RP)| ID | SBP |
|---|---|
| RP5002 | 121 |
| RP5012 | 118 |
| RP5113 | 112 |
| RP5213 | 142 |