GSE146726 Processing Pipeline

Publication

Genome medicine (2020) — PMID 33287884

Dataset

RNA-Seq of circCDYL knockdown (KD) samples in HepG2, J82, and UMUC3 cells and of GRWD1, IGF2BP1, and IGF2BP2 knockdown (KD) samples in J82 and UMUC3 …

1. 1

   Fastq files demultiplexed using Illumina’s bcl2fastq v2.20.0.422

   bcl2fastq v2.20.0.422 GitHub

   $ Bash example

   # Install bcl2fastq (example, actual installation might vary based on system and Illumina's distribution)
   # This tool is typically provided by Illumina and may require specific licensing or access.
   # For example, it might be available via a system package manager or direct download from Illumina.
   
   # Example command for demultiplexing:
   # Replace /path/to/illumina/run/folder with the actual path to your Illumina run directory (containing BCL files).
   # Replace /path/to/output/fastq with the desired directory for the output FASTQ files.
   # Replace /path/to/SampleSheet.csv with the actual path to your SampleSheet.csv file.
   bcl2fastq --runfolder-dir /path/to/illumina/run/folder \
             --output-dir /path/to/output/fastq \
             --sample-sheet /path/to/SampleSheet.csv \
             --no-lane-splitting # Optional: Prevents splitting output by lane if not desired

   Copy View on GitHub
2. 2

   Adapters removed using Trim Galore v0.4.1

   Trim Galore v0.4.1 GitHub

   $ Bash example

   # Install Trim Galore (requires Cutadapt and FastQC)
   # conda install -c bioconda trim-galore
   
   # Define input files (replace with actual file paths)
   # Assuming paired-end reads, which is common for many assays.
   # If single-end, remove the second input file and the --paired flag.
   INPUT_READ1="input_read1.fastq.gz"
   INPUT_READ2="input_read2.fastq.gz"
   
   # Define output directory
   OUTPUT_DIR="trimmed_reads"
   
   # Create output directory if it doesn't exist
   mkdir -p "${OUTPUT_DIR}"
   
   # Run Trim Galore to remove adapters and perform quality trimming
   # --paired: Process paired-end reads
   # --illumina: Detect and remove Illumina adapters (common default)
   # -o: Specify output directory
   # --fastqc: Run FastQC on the trimmed files (often enabled by default or explicitly added)
   trim_galore --version
   trim_galore \
     --paired \
     --illumina \
     --output_dir "${OUTPUT_DIR}" \
     "${INPUT_READ1}" \
     "${INPUT_READ2}"
   

   Copy View on GitHub
3. 3

   Reads mapped using TopHat2 and Bowtie2 (version 2.1.1)

   Bowtie2 v2.1.1 GitHub

   $ Bash example

   # Install TopHat2 and Bowtie2 (if not already installed)
   # conda install -c bioconda tophat2 bowtie2=2.1.1
   
   # Define variables
   # Reference genome and GTF file are placeholders. For eCLIP, hg19 is often used.
   REFERENCE_GENOME_DIR="/path/to/bowtie2_index/hg19" # Directory containing hg19.1.bt2, hg19.2.bt2, etc.
   REFERENCE_GENOME_PREFIX="hg19" # Prefix for the Bowtie2 index files
   GTF_FILE="/path/to/annotation/hg19.gtf" # GTF file for known transcripts (e.g., from UCSC or Ensembl)
   READS_R1="reads_R1.fastq.gz" # Placeholder for forward reads
   READS_R2="reads_R2.fastq.gz" # Placeholder for reverse reads (if paired-end)
   OUTPUT_DIR="tophat_output"
   THREADS=8 # Number of CPU threads
   
   # Create output directory
   mkdir -p "${OUTPUT_DIR}"
   
   # Run TopHat2 for splice-aware alignment using Bowtie2
   # TopHat2 version 2.0.x or later uses Bowtie2 by default.
   # Parameters are based on common eCLIP workflows (e.g., yeolab/eclip).
   #   -o: Output directory
   #   -p: Number of threads
   #   --library-type fr-firststrand: Specify library type as first-strand (common for eCLIP/RNA-seq)
   #   -G: GTF file for known transcripts (improves alignment accuracy for known junctions)
   #   <bowtie2_index_prefix>: The prefix for the Bowtie2 index files
   #   <reads_R1> [<reads_R2>]: Input FASTQ files
   tophat2 \
     -o "${OUTPUT_DIR}" \
     -p "${THREADS}" \
     --library-type fr-firststrand \
     -G "${GTF_FILE}" \
     "${REFERENCE_GENOME_DIR}/${REFERENCE_GENOME_PREFIX}" \
     "${READS_R1}" "${READS_R2}"
   
   # The output will be in the 'tophat_output' directory, including 'accepted_hits.bam'
   # which is the aligned BAM file.

   Copy View on GitHub
4. 4

   Read counts were estimated using HTSeq (v0.6.1p1)

   HTSeq v0.6.1p GitHub

   $ Bash example

   # Install HTSeq (if not already installed)
   # conda install -c bioconda htseq
   
   # Define input and output files
   INPUT_BAM="aligned_reads.bam" # Replace with your input alignment file (BAM format)
   GENE_ANNOTATION_GTF="Homo_sapiens.GRCh38.109.gtf" # Replace with your gene annotation file (GTF format, e.g., from Ensembl or GENCODE)
   OUTPUT_COUNTS_FILE="gene_read_counts.txt"
   
   # Estimate read counts using htseq-count
   # -f bam: Input file format is BAM
   # -r pos: Input alignment file is sorted by position
   # -s reverse: Strandedness of the library (can be 'no', 'yes', or 'reverse' - adjust as per your library prep)
   # -a 10: Minimum alignment quality score (default is 10)
   # -m union: Mode to handle reads overlapping multiple features (default is 'union')
   htseq-count \
     -f bam \
     -r pos \
     -s reverse \
     -a 10 \
     -m union \
     "${INPUT_BAM}" \
     "${GENE_ANNOTATION_GTF}" \
     > "${OUTPUT_COUNTS_FILE}"

   Copy View on GitHub

Tools Used