GSE72500 Processing Pipeline

Publication

Science (New York, N.Y.) (2015) — PMID 26382853

Dataset

Ro60 iCLIP

1. 1

   Illumina software used for basecalling.

   bcl2fastq (Inferred with models/gemini-2.5-flash) v2.20 (Inferred with models/gemini-2.5-flash) GitHub

   $ Bash example

   # Install bcl2fastq (example using conda)
   # conda install -c bioconda bcl2fastq2
   
   # Define input and output directories
   RUN_FOLDER_DIR="/path/to/illumina/run/folder"
   OUTPUT_DIR="/path/to/output/fastq"
   SAMPLE_SHEET="/path/to/sample_sheet.csv" # Optional, but highly recommended for demultiplexing
   
   # Execute bcl2fastq for basecalling and demultiplexing
   bcl2fastq --runfolder-dir "${RUN_FOLDER_DIR}" \
                     --output-dir "${OUTPUT_DIR}" \
                     --sample-sheet "${SAMPLE_SHEET}" \
                     --no-lane-splitting # Example common parameter, adjust as needed

   Copy View on GitHub
2. 2

   Reads were mapped to human genome build hg19 using STAR (https://code.google.com/p/rna-star/) with the "outFilterMultimapNmax 20" option, then PCR duplicates were removed using unique nmers in the barcode sequence.

   STAR v2.4.x (Inferred with models/gemini-2.5-flash) GitHub

   $ Bash example

   # Install STAR (if not already installed)
   # conda install -c bioconda star
   
   # Define variables
   GENOME_DIR="/path/to/STAR_index/hg19" # Path to the STAR genome index for hg19
   READS="input.fastq.gz" # Input FASTQ file (assuming single-end for this example)
   OUTPUT_PREFIX="mapped_reads" # Prefix for output files
   
   # 1. Map reads to human genome build hg19 using STAR
   #    Parameters: "outFilterMultimapNmax 20"
   STAR --genomeDir "${GENOME_DIR}" \
        --readFilesIn "${READS}" \
        --outFileNamePrefix "${OUTPUT_PREFIX}." \
        --outFilterMultimapNmax 20 \
        --outSAMtype BAM SortedByCoordinate \
        --runThreadN 8 # Adjust number of threads as needed
   
   # 2. Remove PCR duplicates using unique nmers in the barcode sequence
   #    This step typically involves UMI (Unique Molecular Identifier) based deduplication.
   #    The exact command depends on how the barcode sequence is incorporated into the reads
   #    (e.g., in the read header, or at the start of the read sequence).
   #    A common tool for this is `umi_tools dedup`.
   
   # Install umi_tools (if not already installed)
   # conda install -c bioconda umi_tools
   
   # Example using umi_tools dedup (assuming UMI is in the read name after a previous extraction step):
   # umi_tools dedup \
   #     --input "${OUTPUT_PREFIX}.Aligned.sortedByCoord.out.bam" \
   #     --output "${OUTPUT_PREFIX}.deduplicated.bam" \
   #     --extract-method=read_id \
   #     --umi-separator=":" \
   #     --log "${OUTPUT_PREFIX}.deduplication.log"
   # If reads are paired-end, add --paired. If UMI needs to be extracted from the read sequence first, use `umi_tools extract` prior to STAR.

   Copy View on GitHub
3. 3

   Peak calling was performed using pyicoclip (http://regulatorygenomics.upf.edu/Software/Pyicoteo/pyicoclip.html) using RefSeq genes as the region file.

   RefSeq vv0.1.1

   $ Bash example

   # Install pyicoteo (which includes pyicoclip)
   # pip install pyicoteo
   
   # Placeholder for input BAM file (aligned reads)
   # Replace with your actual input BAM file
   INPUT_BAM="input.bam"
   
   # Placeholder for RefSeq genes region file (e.g., BED format)
   # This file defines the regions where peaks will be called.
   # Example: Download RefSeq genes for your specific genome assembly (e.g., hg38) 
   # from resources like UCSC Table Browser, Ensembl, or NCBI.
   REFSEQ_GENES_BED="refseq_genes.bed"
   
   # Placeholder for genome FASTA file
   # Replace with your actual genome FASTA file (e.g., hg38.fa)
   GENOME_FASTA="genome.fa"
   
   # Output prefix for peak files
   OUTPUT_PREFIX="pyicoclip_peaks"
   
   # Execute pyicoclip (part of the pyicoteo package)
   pyicoteo clip -i "${INPUT_BAM}" -o "${OUTPUT_PREFIX}" -r "${REFSEQ_GENES_BED}" -g "${GENOME_FASTA}"

   Copy

Tools Used