Upload New File

Codes/HPA_Info_Extracter.R

+# ---------------------------------------------------------
+# HUMAN PROTEIN ATLAS INFO EXTRACTER
+# This R-executable script extracts information from the two downloadable Human Protein Atlas expression datasets at https://www.proteinatlas.org/about/download,
+# for rna (https://www.proteinatlas.org/download/rna_tissue_consensus.tsv.zip) and protein (https://www.proteinatlas.org/download/normal_tissue.tsv.zip) respectively.
+# For each gene, the 3 organs (or tissue groups) with the highest expression in the rna dataset and the protein dataset are identified. In the case of ties,
+# a frequent occurrence in the discrete categories of the protein dataset, more than 3 organs are taken.
+# These data can then be cross-referenced with a list of genes of interest, in our case with the list of genes with CNVs at population level.
+# Before running the script, change to your own reference paths.
+# ---------------------------------------------------------
+
+
+# Load the required library for data manipulation
+library(dplyr)
+
+# ---------------------------------------------------------
+# SECTION 1: GROUP DEFINITION
+# ---------------------------------------------------------
+
+# Define tissue groups for biological categorization
+brain <- c("hippocampal formation", "amygdala", "basal ganglia", "midbrain", "spinal cord", "cerebral cortex", "cerebellum", "hypothalamus", "choroid plexus")
+eye <- c("retina")
+endocrine_tissues <- c("thyroid gland", "parathyroid gland", "adrenal gland", "pituitary gland")
+respiratory_system <- c("nasopharynx", "bronchus", "lung")
+proximal_digestive_tract <- c("oral mucosa", "salivary gland", "esophagus", "tongue")
+liver_and_gallbladder <- c("liver", "gallbladder")
+gastrointestinal_tract <- c("stomach", "small intestine", "colon", "rectum", "duodenum")
+pancreas <- c("pancreas")
+kidney_and_urinary_bladder <- c("kidney", "urinary bladder")
+male_tissues <- c("testis", "epididymis", "prostate", "seminal vesicle")
+female_tissues <- c("endometrium", "vagina", "breast", "cervix", "fallopian tube", "ovary", "placenta")
+muscle_tissues <- c("heart muscle", "skeletal muscle", "smooth muscle")
+connective_and_soft_tissues <- c("soft tissues", "adipose tissue")
+skin <- c("skin")
+bone_marrow_and_lymphoid_tissues <- c("bone marrow", "appendix", "thymus", "spleen", "lymph node", "tonsil")
+
+# Function to map tissues to the defined groups
+map_tissue_to_group <- function(tissue) {
+  if (tissue %in% brain) {
+    return("Brain")
+  } else if (tissue %in% eye) {
+    return("Eye")
+  } else if (tissue %in% endocrine_tissues) {
+    return("Endocrine_tissues")
+  } else if (tissue %in% respiratory_system) {
+    return("Respiratory_system")
+  } else if (tissue %in% proximal_digestive_tract) {
+    return("Proximal_digestive_tract")
+  } else if (tissue %in% liver_and_gallbladder) {
+    return("Liver_and_gallbladder")
+  } else if (tissue %in% gastrointestinal_tract) {
+    return("Gastrointestinal_tract")
+  } else if (tissue %in% pancreas) {
+    return("Pancreas")
+  } else if (tissue %in% kidney_and_urinary_bladder) {
+    return("Kidney_and_urinary_bladder")
+  } else if (tissue %in% male_tissues) {
+    return("Male_tissues")
+  } else if (tissue %in% female_tissues) {
+    return("Female_tissues")
+  } else if (tissue %in% muscle_tissues) {
+    return("Muscle_tissues")
+  } else if (tissue %in% connective_and_soft_tissues) {
+    return("Connective_and_soft_tissues")
+  } else if (tissue %in% skin) {
+    return("Skin")
+  } else if (tissue %in% bone_marrow_and_lymphoid_tissues) {
+    return("Bone_marrow_and_lymphoid_tissues")
+  } else {
+    return("Other")
+  }
+}
+
+# ---------------------------------------------------------
+# SECTION 2: RNA DATA PROCESSING
+# ---------------------------------------------------------
+
+# RNA dataset input: Replace the path with your own RNA dataset path
+rna_expression_data <- read.csv(file = "path/to/rna_tissue_consensus.tsv", sep = "\t")
+
+# Apply tissue group mapping to RNA data
+rna_expression_data$Group <- sapply(rna_expression_data$Tissue, map_tissue_to_group)
+
+# Function to extract the tissue with the highest nTPM (normalized transcripts per million) for each gene
+extract_highest_level_tissue <- function(data) {
+  highest_level_index <- which.max(data$nTPM)
+  highest_level_tissue <- data[highest_level_index, ]
+  return(highest_level_tissue)
+}
+
+# Group RNA data by gene and tissue group, selecting the tissue with the highest nTPM
+all_genes_organs_rna <- rna_expression_data %>%
+  group_by(Gene, Gene.name, Group) %>%
+  filter(nTPM == max(nTPM)) %>%
+  ungroup() %>%
+  as.data.frame()
+
+# Select the top 3 tissues with the highest nTPM for each gene
+all_genes_3organs_rna <- all_genes_organs_rna %>%
+  group_by(Gene) %>%
+  slice_max(order_by = nTPM, n = 3) %>%
+  ungroup() %>%
+  as.data.frame()
+
+# Remove rows where nTPM is equal to 0
+all_genes_3organs_rna <- all_genes_3organs_rna[all_genes_3organs_rna$nTPM != 0, ]
+
+# Export processed RNA data: Replace with your desired output path
+write.table(all_genes_3organs_rna, file = "path/to/output/all_genes_3organs_rna.tsv", sep = "\t", row.names = FALSE, quote = FALSE)
+
+# ---------------------------------------------------------
+# SECTION 3: PROTEIN DATA PROCESSING
+# ---------------------------------------------------------
+
+# Protein dataset input: Replace the path with your own protein dataset path
+protein_expression_data <- read.csv(file = "path/to/normal_tissue.tsv", sep = "\t")
+
+# Apply tissue group mapping to protein data
+protein_expression_data$Group <- sapply(protein_expression_data$Tissue, map_tissue_to_group)
+
+# Remove rows where the reliability of protein expression is uncertain
+protein_expression_data <- protein_expression_data[protein_expression_data$Reliability != "Uncertain", ]
+
+# Function to extract the tissue with the highest protein expression level for each gene
+extract_highest_level_tissue <- function(data) {
+  level_order <- c("N/A", "Ascending", "Descending", "Not representative", "Not detected", "Low", "Medium", "High")
+  highest_level_index <- which.max(match(data$Level, level_order))
+  highest_level_tissue <- data[highest_level_index, ]
+  return(highest_level_tissue)
+}
+
+# Group protein data by gene and tissue group, selecting the tissue with the highest expression level
+all_genes_organs_protein <- protein_expression_data %>%
+  group_by(Gene, Gene.name, Group) %>%
+  do(extract_highest_level_tissue(.)) %>%
+  ungroup()
+
+# Function to extract tissues with the highest levels of protein expression
+extract_highest_levels <- function(data) {
+  level_order <- c("N/A", "Ascending", "Descending", "Not representative", "Low", "Medium", "High")
+  data <- data[data$Level != "Not detected", ]
+  if (nrow(data) == 0) return(data.frame())  # If no data remains, return an empty dataframe
+  highest_level <- max(match(data$Level, level_order))
+  highest_level_tissues <- data[match(data$Level, level_order) == highest_level, ]
+  return(highest_level_tissues)
+}
+
+# Group protein data by gene and extract the top 3 organs with the highest expression levels
+all_genes_3organs_protein <- protein_expression_data %>%
+  group_by(Gene) %>%
+  do(extract_highest_levels(.)) %>%
+  ungroup() %>%
+  as.data.frame()
+
+# Export processed protein data: Replace with your desired output path
+write.table(all_genes_3organs_protein, file = "path/to/output/all_genes_3organs_protein.tsv", sep = "\t", row.names = FALSE, quote = FALSE)
+
+# ---------------------------------------------------------
+# END OF SCRIPT
+# ---------------------------------------------------------