Skip to content

Latest commit

 

History

History
152 lines (104 loc) · 3.53 KB

Dorado.md

File metadata and controls

152 lines (104 loc) · 3.53 KB
Raw

Dorado Basecalling Guide

This guide offers a comprehensive overview of how to set up and run Dorado for ONT sequencing data, specifically focusing on the basecalling step.

Table of Contents

  1. About Dorado
  2. Installation
  3. Usage
    • Simplex Basecalling
    • Duplex Basecalling
  4. Channel-wise Data Splitting with POD5
  5. Running Dorado with SLURM
  6. (Optional) Merging BAM Files

1. About Dorado

(Detailed introduction about Dorado, its advantages, applications, etc.)

2. Installation

 wget https://cdn.oxfordnanoportal.com/software/analysis/dorado-0.3.1-linux-x64.tar.gz
 tar zxf dorado-0.3.1-linux-x64.tar.gz

Note: For optimal performance, Dorado requires POD5 file input. Convert your .fast5 files before basecalling. POD5 File Format

3. Usage

Simplex Basecalling

For running simplex basecalling in Dorado:

#!/bin/bash -l
#SBATCH --account=b1171
#SBATCH --partition=b1171
#SBATCH --gres=gpu:a100:2
#SBATCH --nodes=1
#SBATCH --ntasks=48
#SBATCH --mem=100G
#SBATCH --time=70:00:00

cd $SLURM_SUBMIT_DIR

/home/qgn1237/2_software/dorado-0.3.1-linux-x64/bin/dorado basecaller /home/qgn1237/2_software/dorado-0.3.4-linux-x64/bin/dna_r10.4.1_e8.2_400bps_sup@v4.2.0 sample.pod5 > simplex_dorado.bam

After obtaining the BAM file, extract fastq reads:

samtools view duplex_also_simplex_dorado.bam -d dx:0 | samtools fastq > dorado.simplex.fastq

Duplex Basecalling

For running duplex basecalling:

#!/bin/bash -l
#SBATCH --account=b1171
#SBATCH --partition=b1171
#SBATCH --gres=gpu:a100:2
#SBATCH --nodes=1
#SBATCH --ntasks=48
#SBATCH --mem=100G
#SBATCH --time=70:00:00

cd $SLURM_SUBMIT_DIR

/home/qgn1237/2_software/dorado-0.3.1-linux-x64/bin/dorado duplex /home/qgn1237/2_software/dorado-0.3.4-linux-x64/bin/dna_r10.4.1_e8.2_400bps_sup@v4.2.0 sample.pod5 > duplex_also_simplex_dorado.bam -t 48

After this step, you'll have a BAM file. Separate it into simplex and duplex fastq reads:

samtools view duplex_also_simplex_dorado.bam -d dx:0 | samtools fastq > dorado.simplex.fastq

samtools view duplex_also_simplex_dorado.bam -d dx:1 | samtools fastq > dorado.duplex.fastq
# Or combine all to fastq
samtools view duplex_also_simplex_dorado.bam | samtools fastq > dorado.simplex.fastq

4. Channel-wise Data Splitting with POD5

Firstly, generate the summary.tsv:

pod5 view <path_to_your_pod5_file> --include "read_id, channel" --output summary.tsv

Then, split the data based on channels:

pod5 subset <path_to_your_pod5_file> --summary summary.tsv --columns channel --output <output_directory_path>

5. Running Dorado with SLURM

Create a new SLURM script:

#!/bin/bash -l
#SBATCH --account=b1171
#SBATCH --partition=b1171
#SBATCH --gres=gpu:a100:2
#SBATCH --nodes=1
#SBATCH --ntasks=48
#SBATCH --mem=100G
#SBATCH --time=70:00:00

cd $SLURM_SUBMIT_DIR

for file in <path_to_split_pod5_files>/*.pod5; do
    /home/qgn1237/2_software/dorado-0.3.4-linux-x64/bin/dorado duplex <model_file_path> $file > "${file%.pod5}_dorado.bam"
done

Save the script and submit:

sbatch <name_of_script.sh>

6. (Optional) Merging BAM Files

Merge all BAM files using samtools:

samtools merge merged.bam <path_to_individual_bam_files>/*_dorado.bam

End of Guide

Ensure all software dependencies are installed and paths are correctly specified before running the commands. Always validate the results with the appropriate tools or visualization methods.