lung_CellPhoneDB_normalization.Rmd

```{r}

library(Seurat)
library(readr)
library(tidyverse)

```

```{r}

epi_anno <- readRDS("seurat_objects/epi_anno.RDS")
epi_tumor <- subset(epi_anno, subset = cluster_type == "Tumor" & tissue_type == "Tumor")

imm_anno <- readRDS("seurat_objects/imm_anno.RDS")
imm_subset <- subset(imm_anno, subset = cell_type_imm %in% c("CD14_Macrophages1",
                                                             "CD14_Macrophages2",
                                                             "Myeloid_Dendritic",
                                                             "Plasmacytoid_Dendritic",
                                                             "T_conv1",
                                                             "T_CD8_1",
                                                             "T_CD8_2",
                                                             "NK_cells")
                     & tissue_type == "Tumor")

str_anno <- readRDS("seurat_objects/str_anno.RDS")
str_subset <- subset(str_anno, subset = cell_type_str %in% c("Myofibroblast1",
                                                             "Myofibroblast2")
                     & tissue_type == "Tumor")

```

```{r}

epi_tumor$cell_type <- epi_tumor$cell_type_epi
imm_subset$cell_type <- imm_subset$cell_type_imm
str_subset$cell_type <- str_subset$cell_type_str

subset_combined <- merge(epi_tumor, c(imm_subset, str_subset))

remove(epi_anno)
remove(epi_tumor)
remove(imm_anno)
remove(imm_subset)
remove(str_anno)
remove(str_subset)

```

```{r}

# generating meta file

meta_data <- FetchData(subset_combined, vars = c("cell_type", "patient_id"))
meta_data <- meta_data %>% mutate(pattern = ifelse(patient_id %in% c("p032", "p018", "p019", "p024", "p031", "p033"), "A", "B"))
meta_data$patient_id <- NULL
meta_data <- meta_data %>% unite("cell_type", c("cell_type", "pattern"), sep = "_")
meta_data$cell_id <- rownames(meta_data)

write_csv(meta_data, file = "output/CellPhoneDB/meta_data.csv")

```

```{r}

count_raw <- GetAssayData(object = subset_combined, assay = "RNA", slot = "counts")

#write_csv(as.data.frame(count_raw), file = "output/CellPhoneDB/count_raw.csv")
remove(subset_combined)
#count_raw <- read_csv(file = "output/CellPhoneDB/count_raw.csv")

count_raw1 <- count_raw[,1:20000]
count_raw2 <- count_raw[,20001:38134]

count_norm1 <- apply(count_raw1, 2, function(x) (x/sum(x))*10000)
count_norm2 <- apply(count_raw2, 2, function(x) (x/sum(x))*10000)

count_norm <- cbind(count_norm1, count_norm2)

write.csv(count_norm, "output/CellPhoneDB/count_norm.csv")

write.csv(count_raw, "output/CellPhoneDB/count_raw.csv")

```