GSE132971 Processing Pipeline

Publication

RNA (New York, N.Y.) (2020) — PMID 31624092

Dataset

Direct RNA sequencing enables single-nucleotide m6A detection in endogenous transcript isoforms

1. 1

   Oxford Nanopore's Albacore and Guppy base-callers

   Nanopore sequencing v6.5.7

   $ Bash example

   # Install Guppy (example using conda, ensure you have the appropriate GPU drivers if using GPU acceleration)
   # conda create -n guppy_env python=3.8
   # conda activate guppy_env
   # conda install -c bioconda ont-guppy
   
   # Example command for Guppy base-calling
   # This command processes raw .fast5 files from an input directory and outputs .fastq files.
   # The configuration file specifies the base-calling model, which depends on the flow cell and chemistry used.
   
   # Replace /path/to/raw_fast5_data with the actual directory containing your .fast5 files.
   # Replace /path/to/output_fastq_data with your desired output directory for .fastq files.
   # Replace dna_r9.4.1_450bps_hac.cfg with the appropriate configuration file for your specific Nanopore chemistry and flow cell.
   # The --device auto flag attempts to automatically detect and use available GPUs or fall back to CPU.
   
   guppy_basecaller \
     -i /path/to/raw_fast5_data \
     -s /path/to/output_fastq_data \
     -c dna_r9.4.1_450bps_hac.cfg \
     --recursive \
     --fast5_out \
     --compress_fastq \
     --device auto

   Copy
2. 2

   Reads were aligned and resquiggled with Tombo v1.4

   Tombo v1.4 GitHub

   $ Bash example

   # Install Tombo (if not already installed)
   # conda create -n tombo_env python=3.7
   # conda activate tombo_env
   # pip install tombo
   
   # Define variables (replace with actual paths)
   FAST5_DIR="path/to/your/fast5_files"
   GENOME_FASTA="path/to/your/reference/genome/hg38.fa" # Placeholder: latest assembly (e.g., hg38)
   OUTPUT_DIR="path/to/output/resquiggle_files"
   NUM_PROCESSES=8 # Example number of processes
   
   # Create output directory if it doesn't exist
   mkdir -p "${OUTPUT_DIR}"
   
   # Align reads and resquiggle with Tombo
   tombo resquiggle \
       --fast5-basedirs "${FAST5_DIR}" \
       --genome-fasta "${GENOME_FASTA}" \
       --output-path "${OUTPUT_DIR}/resquiggle_results.hdf5" \
       --processes "${NUM_PROCESSES}" \
       --overwrite

   Copy View on GitHub
3. 3

   Coverage and fraction modified files were obtained using Tombo's text_output commands.

   Tombo v1.5.1 (Inferred with models/gemini-2.5-flash) GitHub

   $ Bash example

   # Install Tombo (if not already installed)
   # conda install -c bioconda tombo
   # pip install tombo
   
   # Placeholder for input Tombo HDF5 file, typically generated by 'tombo detect_modifications'
   TOMBO_HDF5_FILE="sample_modifications.tombo.hdf5"
   
   # Obtain coverage file
   tombo text_output coverage \
       --file-type tombo_stats \
       --output-filepath sample_coverage.txt \
       "${TOMBO_HDF5_FILE}"
   
   # Obtain fraction modified file
   tombo text_output fraction_modified \
       --file-type tombo_stats \
       --output-filepath sample_fraction_modified.txt \
       "${TOMBO_HDF5_FILE}"

   Copy View on GitHub
4. 4

   Bedobs was used to convert the wig files to bed files before being interescted with the coverage file using bedtools.

   bedtools v2.30.0 GitHub

   $ Bash example

   # Install bedtools (example)
   # conda install -c bioconda bedtools
   
   # Bedobs is mentioned as a tool to convert wig files to bed files.
   # Assuming 'Bedobs' is an executable or script available in the environment.
   # If 'Bedobs' is not a standard tool, this step might use a custom script or another tool like UCSC wig2bed.
   # Example for converting a single wig file:
   Bedobs input.wig > converted.bed
   
   # Intersect the converted bed file with the coverage file using bedtools
   # Assuming coverage.bed is the coverage file
   bedtools intersect -a converted.bed -b coverage.bed > intersected_output.bed

   Copy View on GitHub

Tools Used