Environmental-Adaptations-and-Temperature-Effects-on-C.-elegans-Strain-Distribution

First, download strain data from: https://caendr.org/data/data-release/c-elegans/latest The csv file name is: 20231213_c_elegans_strain_data This dataset Includes strain, isotype, location information, and more.

Follow the R script code attached and these are the figures you can expect to get once done with this analysis!

This code was created on macOS 15.3

Install necessary packages if not already installed

install.packages("tidyverse") install.packages("BiocManager") BiocManager::install("DESeq2") install.packages("readr") install.packages("pheatmap") install.packages("ggplot2") install.packages("dplyr") install.packages("factoextra")

Load packages

library(tidyverse) library(DESeq2) library(pheatmap) library(ggplot2) library(BiocManager) library(readr) library(readr) library(factoextra)

#File Directory getwd() setwd("/Users/lesliesouthern/Desktop/caenor")

Change this to your actual folder path

df <- read_csv("strain_data.csv")

Read the CSV file

df <- read.csv("strain_data.csv", stringsAsFactors = FALSE)

Read the CSV with automatic column type detection

df <- read_csv("strain_data.csv", show_col_types = FALSE)

Read the CSV, skipping the first row (which is acting as column names)

df <- read_csv("strain_data.csv", col_names = TRUE, skip = 1, show_col_types = FALSE)

Check the first few rows and structure

head(df) str(df) colnames(df)

sapply(df, function(x) length(unique(x))) # Count unique values per column

ggplot(df, aes(x = as.numeric(ambient_temp))) + geom_histogram(bins = 30, fill = "pink", alpha = 0.6) + labs(title = "Distribution of Ambient Temperatures", x = "Temperature (°C)", y = "Frequency")

ggplot(df, aes(x = landscape, fill = landscape)) + geom_bar() + theme(axis.text.x = element_text(angle = 90, hjust = 1)) + labs(title = "Strain Distribution by Landscape", x = "Landscape Type", y = "Count")

ggplot(df, aes(x = substrate, y = as.numeric(ambient_temp))) + geom_boxplot() + theme(axis.text.x = element_text(angle = 90, hjust = 1)) + labs(title = "Temperature Variation by Substrate Type", x = "Substrate Type", y = "Temperature (°C)")

anova_result <- aov(as.numeric(ambient_temp) ~ landscape, data = df) summary(anova_result)

chisq.test(table(df$substrate, df$landscape))

• Since the p-value is very small, it means there is a significant association between substrate type and landscape.
• In other words, C. elegans strains are not randomly distributed across landscapes based on their substrate preferences.

Select relevant numerical columns

df_numeric <- df[, c("latitude", "longitude", "ambient_temp", "ambient_humidity", "substrate_temp")]

Convert to numeric (ignoring warnings)

df_numeric <- data.frame(lapply(df_numeric, function(x) as.numeric(as.character(x))))

Remove rows with missing values

df_numeric <- na.omit(df_numeric)

Perform PCA

pca_result <- prcomp(df_numeric, scale. = TRUE)

View summary of PCA

summary(pca_result) library(ggplot2)

PC1 (53.99%) explains most of the variation in your data. PC2 (30.26%) explains additional variation. Together, PC1 + PC2 = 84.25%, meaning the first two PCs capture most of the meaningful differences.

Extract PCA loadings (importance of each trait)

pca_loadings <- pca_result$rotation

View loadings for PC1 and PC2

pca_loadings[, 1:2]

PC1 (53.99% Variance) Latitude & Longitude drive variation → likely reflecting geographic structure in strains. Higher temperatures increase PC1 → suggesting temperature adaptation. PC2 (30.26% Variance) Substrate Temp & Humidity are major drivers → indicating local microenvironment adaptations.

Load required libraries

library(ggplot2) library(cluster) library(dplyr) library(tidyverse)

str(df) # See the structure of your original dataframe class(env_data) # Should return "data.frame" head(env_data) colSums(is.na(df)) df$ambient_temp[is.na(df$ambient_temp)] <- mean(df$ambient_temp, na.rm = TRUE) env_data <- df %>% select(latitude, longitude, ambient_temp, ambient_humidity, substrate_temp) %>% drop_na() # Removes rows where any of these are still NA set.seed(123) km_res <- kmeans(scale(env_data), centers = 3, nstart = 25) library(ggplot2) env_data$cluster <- as.factor(km_res$cluster)

ggplot(env_data, aes(x = latitude, y = longitude, color = cluster)) + geom_point(size = 3) + theme_minimal() + labs(title = "Clustering of C. elegans Strains by Environmental Adaptation")