GSE176045 Processing Pipeline

Publication

Nature communications (2021) — PMID 34732726

Dataset

RNA-Seq of isogenic human iPS cell-derived cardiomyocytes with RBM20 mutations created by genome editing (WTB RNA-Seq)

1. 1

   RNA-Seq data was aligned to the human hg19 reference genome and transcriptome (Ensembl 72) using the software STAR.

   STAR v2.4.2a GitHub

   $ Bash example

   # Install STAR (if not already installed)
   # conda install -c bioconda star=2.4.2a
   
   # Create a directory for reference files
   mkdir -p reference_data
   cd reference_data
   
   # Download human hg19 reference genome (UCSC hg19)
   wget -O hg19.fa.gz http://hgdownload.cse.ucsc.edu/goldenPath/hg19/bigZips/hg19.fa.gz
   gunzip hg19.fa.gz
   
   # Download Ensembl 72 GTF annotation for GRCh37 (hg19)
   wget -O Homo_sapiens.GRCh37.72.gtf.gz ftp://ftp.ensembl.org/pub/release-72/gtf/homo_sapiens/Homo_sapiens.GRCh37.72.gtf.gz
   gunzip Homo_sapiens.GRCh37.72.gtf.gz
   
   cd ..
   
   # Create a directory for the STAR genome index
   mkdir -p STAR_index_hg19_Ensembl72
   
   # Build the STAR genome index using hg19 and Ensembl 72 GTF
   STAR --runMode genomeGenerate \
        --genomeDir STAR_index_hg19_Ensembl72 \
        --genomeFastaFiles reference_data/hg19.fa \
        --sjdbGTFfile reference_data/Homo_sapiens.GRCh37.72.gtf \
        --sjdbOverhang 100 \
        --runThreadN 8 # Adjust thread count as needed
   
   # Assuming input RNA-Seq FASTQ files are named read1.fastq.gz and read2.fastq.gz
   # Replace with your actual input file names
   
   # Perform RNA-Seq alignment using STAR
   STAR --runMode alignReads \
        --genomeDir STAR_index_hg19_Ensembl72 \
        --readFilesIn read1.fastq.gz read2.fastq.gz \
        --readFilesCommand zcat \
        --outFileNamePrefix aligned_reads_ \
        --outSAMtype BAM SortedByCoordinate \
        --outSAMattributes Standard \
        --quantMode GeneCounts \
        --twopassMode Basic \
        --runThreadN 8 # Adjust thread count as needed
   

   Copy View on GitHub
2. 2

   To accurately detect diverse alternative splicing events in iPSC-CM RNA-Seq, we adapted a recently developed Percent Spliced In (PSI) splicing method called MultiPath-PSI.

   RNA-seq v1.0.0 (Inferred with models/gemini-2.5-flash)

   $ Bash example

   # Install MultiPath-PSI (example, check official documentation for latest instructions)
   # git clone https://github.com/zhanglab-ucsf/MultiPath-PSI.git
   # cd MultiPath-PSI
   # pip install .
   
   # Define input files and reference data placeholders
   # Replace with actual paths to your iPSC-CM RNA-Seq BAM files, GTF annotation, and genome FASTA.
   INPUT_BAM="path/to/your/ipsc_cm_rnaseq_sample.bam"
   GTF_FILE="path/to/your/gencode.v38.annotation.gtf" # Example: Gencode v38 for human hg38
   GENOME_FASTA="path/to/your/GRCh38.primary_assembly.genome.fa" # Example: GRCh38 primary assembly
   OUTPUT_DIR="multipath_psi_output"
   
   # Create output directory if it doesn't exist
   mkdir -p "${OUTPUT_DIR}"
   
   # Run MultiPath-PSI to accurately detect diverse alternative splicing events
   # This command is based on the typical usage pattern from the MultiPath-PSI GitHub repository.
   # Adjust parameters like --threads, --min_reads, etc., as needed for your specific analysis.
   multipath_psi run \
       --bam_file "${INPUT_BAM}" \
       --gtf_file "${GTF_FILE}" \
       --genome_fasta "${GENOME_FASTA}" \
       --output_dir "${OUTPUT_DIR}" \
       --threads 8 # Example: Use 8 CPU threads for processing

   Copy
3. 3

   MultiPath-PSI is available through AltAnalyze version 2.1.1, requiring aligned BAM files as input and was extensively benchmarked against other local splicing variation approaches (http://altanalyze.readthedocs.io/en/latest/Algorithms/#multipath-psi-splicing-algorithm).

   AltAnalyze v2.1.1 GitHub

   $ Bash example

   # Installation: AltAnalyze can be installed via pip or downloaded from its website/GitHub.
   # To install a specific version (e.g., 2.1.1), it might be necessary to download the source
   # or use a specific release tag if available via pip.
   # pip install AltAnalyze==2.1.1 # (If this specific version is available via pip)
   # Alternatively, clone the repository and run from source:
   # git clone https://github.com/altanalyze/altanalyze.git
   # cd altanalyze
   # python setup.py install # Or just run AltAnalyze.py directly
   
   # Define input and output directories
   INPUT_BAM_DIR="/path/to/aligned_bam_files" # Directory containing aligned BAM files
   OUTPUT_DIR="/path/to/multipath_psi_results" # Directory for AltAnalyze output
   
   # AltAnalyze typically requires a grouping file for differential splicing analysis.
   # Create a placeholder grouping file (e.g., RNASeq_groups.txt) if not provided.
   # Format: SampleID (without .bam extension)	Group	Batch (tab-separated)
   # Example:
   # echo -e "sample1\tControl\tBatch1\nsample2\tControl\tBatch1\nsample3\tTreated\tBatch1\nsample4\tTreated\tBatch1" > RNASeq_groups.txt
   GROUPING_FILE="RNASeq_groups.txt"
   # Create a dummy grouping file for demonstration if it doesn't exist
   if [ ! -f "${GROUPING_FILE}" ]; then
       echo -e "sample1\tControl\tBatch1\nsample2\tControl\tBatch1" > "${GROUPING_FILE}"
       echo "# NOTE: A proper grouping file with all sample IDs from your BAMs is required for meaningful analysis." >> "${GROUPING_FILE}"
   fi
   
   # Execute AltAnalyze with MultiPath-PSI algorithm
   # --species: Reference species (e.g., Hs for Homo sapiens, Mm for Mus musculus).
   # --platform RNASeq: Specifies RNA-Seq data as input.
   # --inputdir: Path to the directory containing input BAM files.
   # --outputdir: Path to the directory where results will be saved.
   # --runMultiPathPSI yes: Activates the MultiPath-PSI algorithm.
   # --grouping: Path to the grouping file for differential analysis.
   python AltAnalyze.py \
       --species Hs \
       --platform RNASeq \
       --inputdir "${INPUT_BAM_DIR}" \
       --outputdir "${OUTPUT_DIR}" \
       --runMultiPathPSI yes \
       --grouping "${GROUPING_FILE}"

   Copy View on GitHub
4. 4

   For gene expression Quantification Kallisto (https://pachterlab.github.io/kallisto/) was used.

   kallisto v0.46.2 GitHub

   $ Bash example

   # Install kallisto (example using conda)
   # conda create -n kallisto_env kallisto=0.46.2 -c bioconda -c conda-forge
   # conda activate kallisto_env
   
   # Placeholder for reference transcriptome index
   # This index should be built once from a FASTA file of transcripts (e.g., from GENCODE or Ensembl).
   # Example command to build index (assuming human_gencode_vXX_transcripts.fasta.gz is available):
   # kallisto index -i human_gencode_vXX_transcriptome.idx human_gencode_vXX_transcripts.fasta.gz
   
   # Define variables
   TRANSCRIPTOME_INDEX="human_gencode_vXX_transcriptome.idx" # Placeholder for latest human GENCODE transcriptome index
   READS_R1="sample_R1.fastq.gz" # Placeholder for input forward reads
   READS_R2="sample_R2.fastq.gz" # Placeholder for input reverse reads
   OUTPUT_DIR="kallisto_quant_output"
   THREADS=8 # Number of threads to use
   
   # Create output directory if it doesn't exist
   mkdir -p "${OUTPUT_DIR}"
   
   # Run kallisto quantification for paired-end reads
   kallisto quant \
     -i "${TRANSCRIPTOME_INDEX}" \
     -o "${OUTPUT_DIR}" \
     -t "${THREADS}" \
     --bias \
     "${READS_R1}" \
     "${READS_R2}"

   Copy View on GitHub

Tools Used