scRNA_GSM3215435.Rmd

---
title: "scRNA-seq analisys Kazantseva"
output: html_document
---


Preprocessing: filtering out bad cells and normalization
UMAP + clustering
Marker selection for clusters
GSM3215435



```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
if (!requireNamespace("Seurat", quietly = TRUE)) install.packages("Seurat")
if (!requireNamespace("ggplot2", quietly = TRUE)) install.packages("ggplot2")
if (!requireNamespace("MAST", quietly = TRUE)) BiocManager::install("MAST")
if (!requireNamespace("dplyr", quietly = TRUE)) install.packages("dplyr")
if (!requireNamespace("Matrix", quietly = TRUE)) install.packages("Matrix")

library(Seurat)
library(Matrix)
library(MAST)
library(ggplot2)
library(dplyr)

```

## R Markdown


```{r}
data <- Read10X("Downloads/GSM3215435/")
dim(data)


plotData <- data.frame(
  umis <- colSums(data)
)

ggplot(data=plotData, aes(x=umis)) +
  geom_histogram() + theme_bw()

```


```{r}
seurat <- CreateSeuratObject(data, min.cells = 10, min.features = 10)
dim(seurat)
```
```{r,message=FALSE}

seurat[["percent.mt"]] <- PercentageFeatureSet(seurat, pattern = "^MT-")
FeatureScatter(seurat, "nCount_RNA", "nFeature_RNA") + scale_x_log10() + scale_y_log10()
FeatureScatter(seurat, "nCount_RNA", "percent.mt")+scale_x_log10()
FeatureScatter(seurat, "nFeature_RNA", "percent.mt") + scale_x_log10()



```

```{r,message=FALSE}
seurat <- subset(seurat, subset = nFeature_RNA > 1000)
dim(seurat)




seurat <- SCTransform(seurat, vars.to.regress = "percent.mt", verbose = FALSE)
VariableFeaturePlot(seurat) + scale_y_log10()
top10_variable_genes <- head(VariableFeatures(seurat), 10)
VariableFeaturePlot(seurat) %>%
LabelPoints(points = top10_variable_genes, repel = TRUE) +
scale_y_log10()

seurat <- RunPCA(seurat, verbose = FALSE)



ElbowPlot(seurat, ndims = 50)



seurat <- RunUMAP(seurat, dims=1:20, verbose = FALSE)

DimPlot(seurat, reduction = "umap") + NoLegend()

#seurat <- RunTSNE(seurat, dims=1:10)
#DimPlot(seurat, reduction = "tsne") + NoLegend()

```

```{r,message=FALSE}

seurat <- FindNeighbors(seurat, dims = 1:20, verbose = FALSE)
seurat <- FindClusters(seurat, resolution=0.6, verbose = FALSE)
DimPlot(seurat, reduction = "umap", label = TRUE) + NoLegend()


FeaturePlot(seurat, c("Cd3e",'Flt3'), cols=c("grey", "red"), reduction="umap", ncol=3)


allMarkers <- FindAllMarkers(seurat, max.cells.per.ident = 100, test.use = "MAST", only.pos = T,assay = "RNA", verbose = FALSE)




goodMarkers <- allMarkers %>% group_by(cluster) %>% top_n(n = 1, wt = avg_log2FC) %>% pull(gene)


FeaturePlot(seurat, goodMarkers[6:10], cols=c("grey", "red"), reduction="umap", ncol=3)

VlnPlot(seurat, goodMarkers[6:10], pt.size = 0.1) 


```