How lukas-1/crispr_4sgrna works
Project type and purpose
• This is an R-based bioinformatics pipeline for designing and analyzing CRISPR guide RNA libraries. It comprises:
• A “library compilation” section under R scripts (functions, preparation, CRISPRa/ko/i builds, analysis)
• An “individual experiments” section that processes sequencing runs (PacBio, Nanopore, Illumina)
• An “arrayed screen” analysis section for plate-based CRISPR screens
• A Shiny app for interactive exploration of compiled libraries and results
• renv.lock for reproducible package management
High-level architectural components
• Reference data sources (Bioconductor org.Hs.eg.db/org.Mm.eg.db, dbSNP, Kaviar, GENCODE, GO, FANTOM5, BioMart)
• R pipeline modules
– Library functions (generic utilities)
– Prepare modules (collect annotations, divide genes into chunks, compile TSS, gene lists, essential genes, transcription factors, secretome, etc.)
– Build modules (CRISPRa, CRISPRko, CRISPRi for human and mouse; interfacing with external design tools: GuideScan, CRISPOR, GPP sgRNA designer)
• Sequencing-run processors
– PacBio pipelines
– Nanopore pipelines
– Illumina paired-end 2sg/4sg pipelines
• Arrayed screen analysis (plate QC, score calculations, visualizations)
• Interactive Shiny application (data ingestion, tidying, annotation, UI)
• Environment manager (renv)
Data stores and flows
• Input files: gene lists (Entrez, Ensembl, HGNC), polymorphism data, pre-existing sgRNA libraries, raw sequencing data
• Intermediate files: mapped sgRNA alignments, GuideScan/CRISPOR output, filtered and annotated guide tables, chunked data frames
• Experimental outputs: FASTA/FASTQ consensus reads, barcode counts, QC metrics, summary TSVs/XLSX
• Final outputs: compiled CRISPR library TSVs, plots (heatmaps, volcano, QC), Shiny app data objects
External tools/services
• GuideScan (off‐line or REST)
• CRISPOR (off‐line or REST)
• Broad GPP sgRNA designer API
• Third-party databases: org.Db packages, dbSNP flat files, Kaviar, GENCODE GTF/annotation, GO annotation files
• Sequencing instruments (PacBio, Sequel-II, Nanopore, Illumina)
Key architectural patterns and principles
• Modular folder structure: each major pipeline step in its own directory
• Separation of concerns: annotation vs. library build vs. sequencing analysis vs. visualization
• File-based pipeline, no centralized database
• Reproducibility via renv (isolated package versions)
• Extensibility: easily add new experiment pipelines under “Individual experiments”