Goals: - Describe the dataset - Analyze the relationship between meditation and gene expression - Use plots to illustrate
Dataset Description: This dataset is extracted from an experiment exploring the effect of meditation on specific gene expression. It has been hypothesized that mindulness meditation could alter biochemical processes within the body. The goal of this experiment was to measure the validity of this statement.
Exploring the Dataset
There are multiple methods to use when exploring datasets in R. But before that, we must set the environment by installing the necessary packages. One of which, is the tidyverse package, it is one of the main packages that make working with data easier and more intuitive.
To initiate tidyverse and load the dataset we will execute
library(tidyverse)## -- Attaching packages --------------------------------------- tidyverse 1.3.1 --## v ggplot2 3.3.5 v purrr 0.3.4
## v tibble 3.1.5 v dplyr 1.0.7
## v tidyr 1.1.4 v stringr 1.4.0
## v readr 2.0.2 v forcats 0.5.1## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()Expression <- read.delim("C:/Users/Glow9/DSDocuments/Misk-DSI-2021-01/Learn_R/data/Expression.txt")One method to gain an overview of the dataset, is the glimpse function. To execute the glimpse() function
glimpse(Expression)## Rows: 21
## Columns: 24
## $ Control_RIPK2_1 <dbl> 0.70, 0.39, 1.22, 1.78, 0.59, 1.47, 1.66, 0.76, 1.4~
## $ Control_RIPK2_2 <dbl> 1.43, 1.82, 0.59, 0.47, 1.03, 1.60, 1.48, 1.67, 0.7~
## $ Meditation_RIPK2_1 <dbl> 1.05, 1.69, 1.22, 1.59, 1.20, 1.99, 0.95, 0.53, 1.3~
## $ Meditation_RIPK2_2 <dbl> 0.90, 1.02, 0.41, 0.88, 1.67, 1.82, 1.03, 1.20, 0.8~
## $ Control_COX2_1 <dbl> -0.08, 2.25, 1.10, 2.71, 1.95, 0.81, -0.17, 1.35, 1~
## $ Control_COX2_2 <dbl> 1.62, 1.01, 1.21, 1.68, 0.45, -0.04, 0.95, 0.80, 0.~
## $ Meditation_COX2_1 <dbl> 4.07, 2.33, 0.97, 0.27, 1.58, 1.32, 0.50, -0.46, 1.~
## $ Meditation_COX2_2 <dbl> -0.65, 0.12, 1.21, -0.02, 0.87, 0.49, 0.91, 0.62, 1~
## $ Control_CCR7_1 <dbl> 2.89, 1.44, 1.64, 1.91, 1.65, 1.05, 1.24, 0.73, 1.1~
## $ Control_CCR7_2 <dbl> 1.50, 0.48, 1.88, 1.86, 0.91, 1.49, 1.34, 1.10, 1.5~
## $ Meditation_CCR7_1 <dbl> 1.50, 0.88, 0.71, 0.86, 1.12, 1.49, 1.94, 1.11, 0.5~
## $ Meditation_CCR7_2 <dbl> 0.64, 0.13, 1.75, 0.26, 1.99, -0.02, 1.50, 0.28, 0.~
## $ Control_CXCR1_1 <dbl> 1.87, 0.27, 1.45, 2.11, 1.31, 1.35, 1.32, -0.11, 0.~
## $ Control_CXCR1_2 <dbl> 1.30, 1.16, 1.40, 1.24, 1.67, 2.75, 1.33, 0.80, 1.0~
## $ Meditation_CXCR1_1 <dbl> 1.01, 0.08, 1.25, 1.87, 1.78, 1.17, 1.14, 1.49, 0.2~
## $ Meditation_CXCR1_2 <dbl> 1.54, 2.21, 2.05, 1.15, 0.63, 0.62, 0.49, 1.98, 1.5~
## $ Control_IL.6_1 <dbl> -0.20, 1.39, 0.15, 1.48, 0.70, 1.07, 2.47, 0.40, 2.~
## $ Control_IL.6_2 <dbl> 2.44, 0.87, 0.31, 1.75, 1.07, 0.10, 2.80, 1.64, 0.2~
## $ Meditation_IL.6_1 <dbl> 1.41, 2.73, 0.39, 1.08, 0.47, 0.49, 2.36, 2.47, 1.9~
## $ Meditation_IL.6_2 <dbl> 1.79, 0.91, 0.78, 0.01, 0.67, 0.16, 1.84, 1.33, 0.6~
## $ Control_TNF_1 <dbl> 1.66, 1.83, 1.18, 0.70, 0.65, 0.76, 0.97, 1.08, 1.1~
## $ Control_TNF_2 <dbl> 1.27, 1.17, 1.84, 1.07, 1.26, 2.12, 1.38, 1.25, 0.5~
## $ Meditation_TNF_1 <dbl> 1.14, 1.04, 0.55, 1.06, 1.22, 0.20, 0.75, 1.14, 0.6~
## $ Meditation_TNF_2 <dbl> 1.06, 0.50, 1.08, 1.57, 0.73, 1.73, 2.42, 1.89, -0.~this gives us an idea of the dataset composition.
We can see that the dataset contains 21 rows and 24 columns, and each column denotes a specific gene divided into an intervention and a control group. We also see that the type of variables is “dbl”, or “double”, meaning that the data type is numeric. In general computing language, numerical vectors are referred to as double.
Another method to explore the data is to compute some summary statistics using the summary() function
summary(Expression)## Control_RIPK2_1 Control_RIPK2_2 Meditation_RIPK2_1 Meditation_RIPK2_2
## Min. :-0.030 Min. :0.080 Min. :-0.270 Min. :0.1500
## 1st Qu.: 0.670 1st Qu.:0.760 1st Qu.: 0.975 1st Qu.:0.8650
## Median : 1.110 Median :1.190 Median : 1.265 Median :0.9700
## Mean : 1.064 Mean :1.059 Mean : 1.274 Mean :0.9661
## 3rd Qu.: 1.470 3rd Qu.:1.320 3rd Qu.: 1.685 3rd Qu.:1.1775
## Max. : 1.890 Max. :1.820 Max. : 2.340 Max. :1.8200
## NA's :3 NA's :3
## Control_COX2_1 Control_COX2_2 Meditation_COX2_1 Meditation_COX2_2
## Min. :-0.250 Min. :-0.0400 Min. :-0.460 Min. :-0.6500
## 1st Qu.: 0.810 1st Qu.: 0.6500 1st Qu.: 0.735 1st Qu.: 0.2750
## Median : 1.340 Median : 0.9500 Median : 1.580 Median : 0.4900
## Mean : 1.331 Mean : 0.9624 Mean : 1.502 Mean : 0.6137
## 3rd Qu.: 1.950 3rd Qu.: 1.3500 3rd Qu.: 2.230 3rd Qu.: 0.9200
## Max. : 2.970 Max. : 1.8600 Max. : 4.070 Max. : 1.6800
## NA's :2 NA's :2
## Control_CCR7_1 Control_CCR7_2 Meditation_CCR7_1 Meditation_CCR7_2
## Min. :0.340 Min. :0.210 Min. :0.260 Min. :-0.020
## 1st Qu.:1.050 1st Qu.:0.710 1st Qu.:0.750 1st Qu.: 0.690
## Median :1.350 Median :1.100 Median :1.110 Median : 0.850
## Mean :1.406 Mean :1.123 Mean :1.139 Mean : 1.044
## 3rd Qu.:1.820 3rd Qu.:1.500 3rd Qu.:1.495 3rd Qu.: 1.595
## Max. :2.890 Max. :1.880 Max. :2.280 Max. : 1.990
## NA's :2 NA's :2
## Control_CXCR1_1 Control_CXCR1_2 Meditation_CXCR1_1 Meditation_CXCR1_2
## Min. :-0.110 Min. :-0.040 Min. :-0.250 Min. :0.310
## 1st Qu.: 0.690 1st Qu.: 0.925 1st Qu.: 0.665 1st Qu.:0.770
## Median : 1.070 Median : 1.230 Median : 1.160 Median :1.540
## Mean : 1.139 Mean : 1.171 Mean : 1.117 Mean :1.311
## 3rd Qu.: 1.400 3rd Qu.: 1.335 3rd Qu.: 1.635 3rd Qu.:1.760
## Max. : 2.490 Max. : 2.750 Max. : 2.150 Max. :2.210
## NA's :2 NA's :2 NA's :2 NA's :2
## Control_IL.6_1 Control_IL.6_2 Meditation_IL.6_1 Meditation_IL.6_2
## Min. :-0.9800 Min. :-0.130 Min. :-0.770 Min. :0.010
## 1st Qu.: 0.4000 1st Qu.: 0.320 1st Qu.: 0.480 1st Qu.:0.325
## Median : 0.7000 Median : 0.910 Median : 1.380 Median :0.670
## Mean : 0.8886 Mean : 1.075 Mean : 1.211 Mean :0.870
## 3rd Qu.: 1.4800 3rd Qu.: 1.640 3rd Qu.: 1.950 3rd Qu.:1.440
## Max. : 2.4700 Max. : 2.800 Max. : 2.730 Max. :2.060
## NA's :2 NA's :2
## Control_TNF_1 Control_TNF_2 Meditation_TNF_1 Meditation_TNF_2
## Min. :0.430 Min. :0.230 Min. :0.2000 Min. :-0.4900
## 1st Qu.:0.760 1st Qu.:1.070 1st Qu.:0.6100 1st Qu.: 0.5900
## Median :1.010 Median :1.270 Median :1.0400 Median : 0.7900
## Mean :1.122 Mean :1.303 Mean :0.9247 Mean : 0.9742
## 3rd Qu.:1.500 3rd Qu.:1.570 3rd Qu.:1.1800 3rd Qu.: 1.4200
## Max. :2.320 Max. :2.120 Max. :1.8000 Max. : 2.4200
## NA's :2 NA's :2It created a summary of estimates including the mean, median, IQR for each column.
Note this dataset does not follow the structure of tidy data, where each column has one variable, each row has one observation, and the dataframe has one observational unit.
In order to transform the data from the untidy wide form to the tidy long form, we will use the tidyr package, function pivot_longer:
Expression %>%
pivot_longer(everything()) -> expression_tidy
View(expression_tidy)The view function from the utils package opens the dataframe in a new tap so we get a visual of the data to see the reflected changes.
Now that we have more manageable long data, we can further tidy it by grouping the variables in a way that makes working with the 504 rows we got from pivoting the data longer. To begin, we will create two new columns to add the treatment group and its corresponding values.
expression_tidy$Control_before <- "."
expression_tidy$Control_after <- "."
expression_tidy$Meditation_before <- "."
expression_tidy$Meditation_after <- "."
expression_tidy$Change <- "."
View(expression_tidy)Now we need to fill the columns with values.