search

GoekeLab / xpore

Identification of differential RNA modifications from nanopore direct RNA sequencing
https://xpore.readthedocs.io/
MIT License
131 stars 23 forks source link

eventalignment and bas fast5 file #159

Closed rezarahman12 closed 1 year ago

rezarahman12 commented 1 year ago

Dear xpore team,

I have downloaded xpore paper's dRNA-seq data consisting of fast5 (ftp://ftp.sra.ebi.ac.uk/vol1/run/ERR470/ERR4706254/HEK293T-Mettl3-KO-rep1.tar.gz) and basecalled fastq files (ftp://ftp.sra.ebi.ac.uk/vol1/run/ERR470/ERR4706158/HEK293T-Mettl3-KO-rep1.fastq.gz) from . Then, I ran a job using nanopolish with the below script.

The output is also given below for your kind consideration. I will seek your opinion is it a normal output since I see a lot of bad fast5 file. ... [readdb] indexing /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/fast5_dir/ko-rep1/20181121_1017_21112018_Sho_METTI3_10/fast5/221 [readdb] num reads: 1490210, num reads with path to fast5: 1113644 fast5 files could not be located for 376566 reads [post-run summary] total reads: 981058, unparseable: 0, qc fail: 40158, could not calibrate: 1979, no alignment: 719, bad fast5: 337754

The script of the job is given below:

!/bin/bash

PBS -A UQ-QBI

PBS -l select=1:ncpus=10:mem=50GB

PBS -e /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/nanopol/ko-rep1

PBS -o /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/nanopol/ko-rep1

PBS -l walltime=72:00:00

PBS -N nanopol_ko-rep1

cd $PBS_O_WORKDIR

module load nanopolish/0.14.0

nanopolish index -d /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/fast5_dir/20181121_1017_21112018_Sho_METTI3_10/fast5 /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/fastq_dir/ko-rep1/HEK293T-Mettl3-KO-rep1.fastq.gz nanopolish eventalign --reads /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/fastq_dir/ko-rep1/HEK293T-Mettl3-KO-rep1.fastq.gz --bam /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/align/ko-rep1/aligned.sort.bam --genome /scratch/project_mnt/S0077/xPore/Homo_sapiens.GRCh38.cdna.all.fa --signal-index --scale-events --summary /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/nanopol/ko-rep1/summary.txt --threads 32 > /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/nanopol/ko-rep1/eventalign.txt

I do highly appreciate your time, help and cooperation. Kind regards Reza

yuukiiwa commented 1 year ago

Dear Reza (I will tag you here @rezarahman12),

It will be great if you try gunzip HEK293T-Mettl3-KO-rep1.fastq.gz first before indexing:

gunzip HEK293T-Mettl3-KO-rep1.fastq.gz

nanopolish index -d /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/fast5_dir/20181121_1017_21112018_Sho_METTI3_10/fast5 /scratch/project_mnt/S0077/xPore/rawdata/xpore_data/fastq_dir/ko-rep1/HEK293T-Mettl3-KO-rep1.fastq

Thanks!

Best wishes, Yuk Kei

rezarahman12 commented 1 year ago

Thank you for your kind feedback. @yuukiiwa

rezarahman12 commented 1 year ago

Hi Yuk Kei, (@yuukiiwa ) Thank you the problem resolved. Best regards Reza