Base-editing screen analysis pipeline

A modular R-based pipeline for processing CRISPR gRNA base-editing screening data from raw FASTQ files through alignment, read counting, and statistical analysis (z-scores, p-values, and FDR), using control-guide log fold-changes as the null distribution.

This repository is designed to support reproducible, end-to-end analysis of pooled CRISPR base-editing screens, enabling standardized processing from sequencing reads to statistically inferred guide-level effects.

An accompanying .Rmd file is provided to demonstrate a full example workflow, allowing users to run the pipeline step-by-step on example data and reproduce the analysis from raw sequencing reads to final statistical outputs.

---

Overview

This pipeline performs:

Data Processing : fastq_to_readcounts_pipeline.R

• FASTQ preprocessing (trimming, motif-based filtering)
• gRNA reference index construction
• Alignment to gRNA library
• BAM sorting and indexing
• Read counting per gRNA
• Merging sample-level count matrices

Statistical Analysis : readcounts_to_pvalues.R

• RPM normalization
• Replicate averaging
• Log2 transformation
• Log fold-change calculation
• Outlier filtering (IQR-based)
• Z-score normalization (control-based)
• p-value estimation
• Benjamini–Hochberg FDR correction

---

---

Installation

Install required R packages:

install.packages(c(
  "tidyverse",
  "readxl",
  "openxlsx",
  "stringr",
  "data.table"
))

if (!requireNamespace("BiocManager", quietly = TRUE))
  install.packages("BiocManager")

BiocManager::install(c(
  "Rsubread",
  "Biostrings",
  "GenomicAlignments",
  "GenomicFeatures",
  "QuasR",
  "Rsamtools",
  "ShortRead"
))