---
title: "MAGeCKFlute - Integrative analysis pipeline for pooled CRISPR functional genetic screens"
author: "Binbin Wang, Mei Wang, Wubing Zhang, Wei Li & X. Shirley Liu"
date: "`r Sys.Date()`"
output:
  BiocStyle::html_document
bibliography: library.bib
vignette: >
  %\VignetteIndexEntry{MAGeCKFlute.Rmd}
  %\VignetteEngine{knitr::rmarkdown}
  \usepackage[utf8]{inputenc}
abstract: >
  CRISPR screens represent a promising technology to systematically evaluate gene functions. We have previously developed two algorithms, MAGeCK [@Wei2014] and MAGeCK-VISPR [@Wei2015], to analyze CRISPR screen data. The two algorithms use a negative binomial distribution to model variances of sgRNA read counts and allow users to perform read-count mapping, normalization and QC, as well as to identify positively and negatively selected genes in the screens.
  Here, we developed MAGeCKFlute for accurate identification of gene hits and associated pathways. MAGeCKFlute is distinguished from other currently available tools by its comprehensive pipeline, which contains a series of functions for analyzing CRISPR screen data. This vignette explains how to use MAGeCKFlute to perform quality control (QC), normalization, batch effect removal, gene hit identification and downstream functional enrichment analysis for CRISPR screens.
---


<!-- This is the source document -->

```{r setup}
knitr::opts_chunk$set(tidy=FALSE, cache=TRUE,
                      dev="png", message=FALSE, error=FALSE, warning=TRUE)
```	

**Note:** if you use MAGeCKFlute in published research, please cite:
Binbin Wang, Mei Wang, Wubing Zhang. "Integrative analysis of pooled CRISPR genetic screens using MAGeCKFlute." Nature Protocols (2019), doi: [10.1038/s41596-018-0113-7](https://www.nature.com/articles/s41596-018-0113-7).

## Install and load the required packages
```{r library, eval=TRUE}
if(!"MAGeCKFlute" %in% installed.packages()) BiocManager::install("MAGeCKFlute")
if(!"clusterProfiler" %in% installed.packages()) BiocManager::install("clusterProfiler")
if(!"ggplot2" %in% installed.packages()) BiocManager::install("ggplot2")

library(MAGeCKFlute)
library(clusterProfiler)
library(ggplot2)
```


# Quick start

## Downstream analysis of MAGeCK RRA
The MAGeCK (`mageck test`) uses Robust Rank Aggregation (RRA) for robust identification of CRISPR-screen hits, and outputs the summary results at both sgRNA and gene level. Before performing the downstream analysis, please make sure you have got the gene summary and sgRNA summary results from `mageck test`. MAGeCKFlute incorporates an example datasets [@Pan2018], which study the gene functions in T cell mediate tumor killing by performing CRISPR screens in B16F10 cell lines co-cultured with Pmel-1 T cells. There are three samples in the data, including Pmel1_Input (B16F10 cells without T cell co-culture), Pmel1_Ctrl (B16F10 cells co-cultured with control T cells), and Pmel1 (B16F10 cells co-cultured with antigen specific T cells). We compared `Pmel1` with `Pmel1_Ctrl` using `mageck test`, which identifies genes associated with T cell mediated tumor killing.

### Run the FluteRRA pipeline
MAGeCKFlute processes MAGeCK RRA results ("gene_summary" and "sgrna_summary") with the function FluteRRA, which identifies positively and negatively selected genes and performs functional analysis.
```{r fluterra, eval=FALSE}
## path to the gene summary file (required)
file1 = file.path(system.file("extdata", package = "MAGeCKFlute"),
                  "testdata/rra.gene_summary.txt")
## path to the sgRNA summary file (optional)
file2 = file.path(system.file("extdata", package = "MAGeCKFlute"),
                  "testdata/rra.sgrna_summary.txt")
# Run FluteRRA with only gene summary file
FluteRRA(file1, proj="Pmel1", organism="mmu", outdir = "./")

# Run FluteRRA with both gene summary file and sgRNA summary file
FluteRRA(file1, file2, proj="Pmel1", organism="mmu", outdir = "./")
```

### Important parameters
* `incorporateDepmap` will allow users to compare the selected genes in the input data and Depmap data, which could clean the selected genes by eliminating lethal genes (false positive hits in some CRISPR screens) and then perform downstream enrichment analysis. You can set the `incorporateDepmap` to be `TRUE` if you want to include this analysis. Users could specify Depmap data of specific cell lines by setting the `cell_lines` and `lineages`.

```{r incorporateDepmap, eval=FALSE}
FluteRRA(file1, proj="Pmel1", organism="mmu", incorporateDepmap = TRUE, outdir = "./")
```

* `omitEssential` allows users to eliminate all common essential genes selected in the Depmap from all of the downstream analysis.
```{r omitEssential, eval=FALSE}
FluteRRA(gdata, proj="Pmel1", organism="mmu", omitEssential = TRUE, outdir = "./")
```

**Hints**: all figures and intermediate data are saved into local directory "./MAGeCKFlute_Test/", and all figures are integrated into file "FluteRRA_Test.pdf".

For more available parameters in `FluteRRA`, please read the documentation using the command `?FluteRRA`.


## Downstream analysis of MAGeCK MLE
The MAGeCK-VISPR (`mageck mle`) utilizes a maximum likelihood estimation (MLE) for robust identification of CRISPR screen hits. It outputs **beta scores** and the associated statistics in multiple conditions. The **beta score** describes how a gene is selected: a positive beta score indicates positive selection, and a negative beta score indicates negative selection. Before using `FluteMLE`, you should first get gene summary result from MAGeCK-VISPR (`mageck mle`). MAGeCKFlute uses the same dataset as before for demonstration. Using `mageck mle`, we removed the baseline effect (plasmid sample) from all the three samples, including Pmel1_Input (B16F10 cells without T cell co-culture), Pmel1_Ctrl (B16F10 cells co-cultured with control T cells), and Pmel1 (B16F10 cells co-cultured with antigen specific T cells).

### Run the FluteMLE pipeline
MAGeCKFlute processes MAGeCK MLE results ("gene_summary") with the function FluteMLE, which performs QC and data normalization based on the beta score, identifies essential genes by comparing control and treatment samples, and finally performs functional enrichment analysis.
```{r flutemle, eval=FALSE}
file3 = file.path(system.file("extdata", package = "MAGeCKFlute"),
                  "testdata/mle.gene_summary.txt")
FluteMLE(file3, treatname="Pmel1", ctrlname="Pmel1_Ctrl", proj="Pmel1", organism="mmu")
```


### Important parameters
* `incorporateDepmap` will allow users to compare the selected genes in the input data and Depmap data, which could clean the selected genes by eliminating lethal genes (false positive hits in some CRISPR screens) and then perform downstream enrichment analysis. You can set the `incorporateDepmap` to be `TRUE` if you want to include this analysis. Users could specify Depmap data of specific cell lines by setting the `cell_lines` and `lineages`.

```{r incorporateDepmap2, eval=FALSE}
## Take Depmap screen as control
FluteMLE(gdata, treatname="Pmel1_Ctrl", ctrlname="Depmap", proj="Pmel1", organism="mmu", incorporateDepmap = TRUE)
```

* `omitEssential` allows users to eliminate all common essential genes selected in the Depmap from all of the downstream analysis.
```{r omitEssential2, eval=FALSE}
FluteMLE(gdata, treatname="Pmel1", ctrlname="Pmel1_Ctrl", proj="Pmel1", organism="mmu", omitEssential = TRUE)
```

**Hint**: All pipeline results are written into local directory "./MAGeCKFlute_Test/", and all figures are integrated into file "FluteMLE_Test.pdf".

For more available parameters in `FluteMLE`, please read the documentation using the command `?FluteMLE`.


# Step by step analysis

## Section I: Quality control

### Input data
MAGeCK/MAGeCK-VISPR outputs a count summary file, which summarizes some basic QC scores at raw count level, including map ratio, Gini index, and NegSelQC. MAGeCKFlute incorporates an example datasets [@Ophir2014] for demonstration.

```{r CheckCountSummary}
file4 = file.path(system.file("extdata", package = "MAGeCKFlute"),
                  "testdata/countsummary.txt")
countsummary = read.delim(file4, check.names = FALSE)
head(countsummary)
```

### Visualize the QC results
```{r CountQC, fig.height=5, fig.width=4.5}
# Gini index
BarView(countsummary, x = "Label", y = "GiniIndex",
        ylab = "Gini index", main = "Evenness of sgRNA reads")

# Missed sgRNAs
countsummary$Missed = log10(countsummary$Zerocounts)
BarView(countsummary, x = "Label", y = "Missed", fill = "#394E80",
        ylab = "Log10 missed gRNAs", main = "Missed sgRNAs")

# Read mapping
MapRatesView(countsummary)
# Or
countsummary$Unmapped = countsummary$Reads - countsummary$Mapped
gg = reshape2::melt(countsummary[, c("Label", "Mapped", "Unmapped")], id.vars = "Label")
gg$variable = factor(gg$variable, levels = c("Unmapped", "Mapped"))
gg = gg[order(gg$Label, gg$variable), ]
p = BarView(gg, x = "Label", y = "value", fill = "variable", 
            position = "stack", xlab = NULL, ylab = "Reads", main = "Map ratio")
p + scale_fill_manual(values = c("#9BC7E9", "#1C6DAB"))

```


## Section II: Downstream analysis of MAGeCK RRA

For CRISPR/Cas9 screens with two experimental conditions, MAGeCK-RRA is available for  identification of essential genes. In MAGeCK-RRA results, the sgRNA summary and gene summary file summarizes the statistical significance of positive selections and negative selections at sgRNA level and gene level.

### Read the required data
```{r CheckRRARes}
file1 = file.path(system.file("extdata", package = "MAGeCKFlute"),
                  "testdata/rra.gene_summary.txt")
gdata = ReadRRA(file1)
head(gdata)

file2 = file.path(system.file("extdata", package = "MAGeCKFlute"),
                  "testdata/rra.sgrna_summary.txt")
sdata = ReadsgRRA(file2)
head(sdata)
```

## To incorporate depmap data that are profiled in human cell lines, we will convert mouse gene names to homologous human genes for this dataset.
```{r mmu2hsa}
gdata$HumanGene = TransGeneID(gdata$id, fromType = "symbol", toType = "symbol",
                              fromOrg = "mmu", toOrg = "hsa")
sdata$HumanGene = TransGeneID(sdata$Gene, fromType = "symbol", toType = "symbol",
                              fromOrg = "mmu", toOrg = "hsa")
```

### Compute the similarity between the CRISPR screen with Depmap screens
```{r Depmap, eval=FALSE}
## Remove missing or duplicate human genes
idx = duplicated(gdata$HumanGene)|is.na(gdata$HumanGene)
gdata = gdata[!idx, ]
depmap_similarity = ResembleDepmap(gdata, symbol = "HumanGene", score = "Score")
head(depmap_similarity)
```

### Omit common essential genes from the data
```{r omitessential, eval=FALSE}
gdata = OmitCommonEssential(gdata, symbol = "HumanGene")
sdata = OmitCommonEssential(sdata, symbol = "HumanGene")
```

### Visualization of negative selections and positive selections
#### Volcano plot
```{r selection1, fig.height=4, fig.width=7}
gdata$LogFDR = -log10(gdata$FDR)
p1 = ScatterView(gdata, x = "Score", y = "LogFDR", label = "id", 
                 model = "volcano", top = 5)
print(p1)

# Or
p2 = VolcanoView(gdata, x = "Score", y = "FDR", Label = "id")
print(p2)

```

#### Rank plot
Rank all the genes based on their scores and label genes in the rank plot.
```{r rankrra, fig.height=6, fig.width=4}
gdata$Rank = rank(gdata$Score)
p1 = ScatterView(gdata, x = "Rank", y = "Score", label = "id", 
                 top = 5, auto_cut_y = TRUE, ylab = "Log2FC", 
                 groups = c("top", "bottom"))
print(p1)
```

Label interested hits using parameter `toplabels` (in ScatterView) and `genelist` (in RankView).
```{r, fig.height=4, fig.width=2.5}
ScatterView(gdata, x = "Rank", y = "Score", label = "id", top = 5, 
            auto_cut_y = TRUE, groups = c("top", "bottom"), 
            ylab = "Log2FC", toplabels = c("Pbrm1", "Arid2", "Brd7"))
```

You can also use the function `RankView` to draw the figure.
```{r rankrra2, fig.height=4, fig.width=3}
geneList= gdata$Score
names(geneList) = gdata$id
p2 = RankView(geneList, top = 5, bottom = 10)
print(p2)
RankView(geneList, top = 0, bottom = 0, genelist = c("Pbrm1", "Arid2", "Brd7"))
```

#### Dot plot
Visualize negative and positive selected genes separately.
```{r scatter, fig.height=4, fig.width=6}
gdata$RandomIndex = sample(1:nrow(gdata), nrow(gdata))
gdata = gdata[order(-gdata$Score), ]
gg = gdata[gdata$Score>0, ]
p1 = ScatterView(gg, x = "RandomIndex", y = "Score", label = "id",
                 y_cut = CutoffCalling(gdata$Score,2), 
                 groups = "top", top = 5, ylab = "Log2FC")
p1
gg = gdata[gdata$Score<0, ]
p2 = ScatterView(gg, x = "RandomIndex", y = "Score", label = "id",
                 y_cut = CutoffCalling(gdata$Score,2), 
                 groups = "bottom", top = 5, ylab = "Log2FC")
p2
```


#### sgRankView - visualize the rank of sgRNAs targeting top selected genes.
```{r sgRNARank, fig.height=4, fig.width=7}
p2 = sgRankView(sdata, top = 4, bottom = 4)
print(p2)
```

### Enrichment analysis 
For more information about functional enrichment analysis in MAGeCKFlute, please read the [MAGeCKFlute_enrichment document](https://www.bioconductor.org/packages/3.11/bioc/vignettes/MAGeCKFlute/inst/doc/MAGeCKFlute_enrichment.html), in which we introduce all the available options and methods.
```{r enrich_rra, fig.height=4, fig.width=9}
geneList= gdata$Score
names(geneList) = gdata$id
enrich_pos = EnrichAnalyzer(geneList = geneList[geneList>0.5], 
                            method = "HGT", type = "KEGG")
enrich_neg = EnrichAnalyzer(geneList = geneList[geneList< -0.5], 
                            method = "HGT", type = "KEGG")
```

#### Visualization of enrichment results
```{r enrichview}
EnrichedView(enrich_pos, mode = 1, top = 5, bottom = 0)
EnrichedView(enrich_pos, mode = 2, top = 5, bottom = 0)
EnrichedView(enrich_neg, mode = 2, top = 0, bottom = 5)
```


## Section III: Downstream analysis of MAGeCK MLE
The MAGeCK-VISPR (`mageck mle`) computes beta scores and the associated statistics for all genes in  multiple conditions. The **beta score** describes how the gene is selected: a positive beta score indicates a positive selection, and a negative beta score indicates a negative selection. Before using `FluteMLE`, you should first get gene summary result from MAGeCK-VISPR (`mageck mle`). MAGeCKFlute incorporates an example datasets for demonstration.

### Batch effect removal (not recommended)

Is there batch effects? This is a commonly asked question before perform later analysis. In our package, we provide `HeatmapView` to ensure whether the batch effect exists in data and use `BatchRemove` to remove easily if same batch samples cluster together.

```{r BatchRemove, fig.height=6, fig.width=9}
##Before batch removal
edata = matrix(c(rnorm(2000, 5), rnorm(2000, 8)), 1000)
colnames(edata) = paste0("s", 1:4)
HeatmapView(cor(edata))

## After batch removal
batchMat = data.frame(sample = colnames(edata), batch = rep(1:2, each = 2))
edata1 = BatchRemove(edata, batchMat)
head(edata1$data)
print(edata1$p)
```

### read gene summary file
```{r ReadBeta}
file3 = file.path(system.file("extdata", package = "MAGeCKFlute"),
                  "testdata/mle.gene_summary.txt")
# Read and visualize the file format
gdata = ReadBeta(file3)
head(gdata)
```

### Normalization of beta scores

It is difficult to control all samples with a consistent cell cycle in a CRISPR screen experiment with multi conditions. Besides, beta score among different states with an inconsistent cell cycle is incomparable. So it is necessary to do the normalization when comparing the beta scores in different conditions. Essential genes are those genes that are indispensable for its survival. The effect generated by knocking out these genes in different cell types is consistent. Based on this, we developed the cell cycle normalization method to shorten the gap of the cell cycle in different conditions.

```{r NormalizeBeta}
ctrlname = "Pmel1_Ctrl"
treatname = "Pmel1"
gdata$HumanGene = TransGeneID(gdata$Gene, fromType = "symbol", toType = "symbol",
                              fromOrg = "mmu", toOrg = "hsa")

gdata_cc = NormalizeBeta(gdata, id = "HumanGene", samples=c(ctrlname, treatname), 
                         method="cell_cycle")
head(gdata_cc)
```


### Distribution of all gene beta scores

After normalization, the distribution of beta scores in different
conditions should be similar. We can evaluate the distribution of beta
scores using the function ‘DensityView’, and ‘ConsistencyView’.

```{r DistributeBeta, fig.height=5, fig.width=8}
DensityView(gdata_cc, samples=c(ctrlname, treatname))
ConsistencyView(gdata_cc, ctrlname, treatname)

# Another option MAView
MAView(gdata_cc, ctrlname, treatname)
```

### Positive selection and negative selection
```{r selection2, fig.height=4, fig.width=5}
gdata_cc$Control = rowMeans(gdata_cc[,ctrlname, drop = FALSE])
gdata_cc$Treatment = rowMeans(gdata_cc[,treatname, drop = FALSE])

p1 = ScatterView(gdata_cc, "Control", "Treatment", label = "Gene", 
                 auto_cut_diag = TRUE, display_cut = TRUE, 
                 groups = c("top", "bottom"),
                 toplabels = c("Pbrm1", "Brd7", "Arid2", "Jak1", "Stat1", "B2m"))
print(p1)
```

#### Rank plot - label top hits
```{r rank, fig.height=4, fig.width=3}
gdata_cc$Diff = gdata_cc$Treatment - gdata_cc$Control
gdata_cc$Rank = rank(gdata_cc$Diff)
p1 = ScatterView(gdata_cc, x = "Rank", y =  "Diff", label = "Gene", 
                 top = 5, auto_cut_y = TRUE, groups = c("top", "bottom"))
print(p1)

# Or
rankdata = gdata_cc$Treatment - gdata_cc$Control
names(rankdata) = gdata_cc$Gene
RankView(rankdata)
```

#### Nine-square scatter plot - identify treatment-associated genes
```{r Square, fig.height=4, fig.width=5}
p1 = ScatterView(gdata_cc, x="Pmel1_Ctrl", y="Pmel1", label = "Gene", 
                 model = "ninesquare", top = 5, display_cut = TRUE, force = 2)
print(p1)
```

Customize the cutoff
```{r Square2, fig.height=4, fig.width=5}
p1 = ScatterView(gdata_cc, x="Pmel1_Ctrl", y="Pmel1", label = "Gene", 
                 model = "ninesquare", top = 5, display_cut = TRUE, 
                 x_cut = c(-1,1), y_cut = c(-1,1))
print(p1)
```

Or
```{r, fig.height=4, fig.width=5}
p2 = SquareView(gdata_cc, label = "Gene", 
                x_cutoff = CutoffCalling(gdata_cc$Control, 2), 
                y_cutoff = CutoffCalling(gdata_cc$Treatment, 2))
print(p2)
```

### Functional analysis for treatment-associated genes
```{r EnrichSquare, fig.height=4, fig.width=6}
# 9-square groups
Square9 = p1$data
idx=Square9$group=="topcenter"
geneList = Square9$Diff[idx]
names(geneList) = Square9$Gene[idx]
universe = Square9$Gene

# Enrichment analysis
kegg1 = EnrichAnalyzer(geneList = geneList, universe = universe)
EnrichedView(kegg1, top = 6, bottom = 0)
```

Also, pathway visualization can be done using function `KeggPathwayView` [@Luo2013].
```{r pathview2, eval=TRUE}
gdata_cc$Entrez = TransGeneID(gdata_cc$Gene, "symbol", "entrez", organism = "mmu")
genedata = gdata_cc[, c("Entrez", "Control","Treatment")]
arrangePathview(genedata, pathways = "mmu04630", organism = "mmu", sub = NULL)
```


# Session info

```{r sessionInfo}
sessionInfo()
```

# References