Empty diffmod.table file, only header generated when running xpore diffmod with preprocessed files, not able to reproduce the data

[kwonej0617]

Dear @ploy-np and @yuukiiwa, @jonathangoeke,

I downloaded preprocessed data of HEK293T-WT/KO-rep1,2,3 from https://doi.org/10.5281/zenodo.4587661 and run xpore 2.1 with those files. However, I didn't get any sites in the output file, diffmod.table, and only the header was printed. Could you please check if I made a mistake? Also, please give me suggestions to generate the result. Thank you so much!

Here is the command line to run xPore diffmod.

module load xpore/2.1
xpore diffmod --config config.yml --n_processes 8

onfig.yml file looks like the following.

Here are the downloaded preprocessed files.

[yuukiiwa]

Hi @kwonej0617,

One of the files from zenodo (HEK293T-WT-0-rep2) is corrupted. It will be great if you can try out the file below:
https://drive.google.com/file/d/1BPWX8x_tJR2BCsMM-fWvSjqqspgJKkZ4/view?usp=sharing

Thanks!

Best wishes,
Yuk Kei

[kwonej0617]

Hi @yuukiiwa.

Thank you for your reply.
Actually, HEK293T-WT-0-rep2 is not the dataset I downloaded to run xPore in this run.
Here are the files I downloaded. Could you please check if any of these files are also corrupted or have some problem?
HEK293T-WT-rep1.tar.gz
HEK293T-WT-rep2.tar.gz
HEK293T-WT-rep3.tar.gz
HEK293T-KO-rep1.tar.gz
HEK293T-KO-rep2.tar.gz
HEK293T-KO-rep3.tar.gz

Thanks!!

[yuukiiwa]

Hi @kwonej0617,

Sorry for the delayed reply! I have found out why you got the empty diffmod.table as the KO datapreps have transcript versions while the WT datapreps don't have transcript versions. Here I am attaching the datapreps for WT, which have transcript versions:
https://drive.google.com/drive/folders/1wUZVtBPV8epxosuXtwiix5Hp6dPe-Z_U?usp=share_link

Thanks!

Best wishes,
Yuk Kei

[kwonej0617]

Hi @yuukiiwa

I really appreciate your reply!
When I decompressed WT1.tar.gz file, it contains data.index, data.json (Usually, when I run xpore dataprep, I got four types of file - data.index, data.json, data.readcount, and eventalign.index). I think xpore diffmod requires those four files ready. Therefore, would it be okay to substitute data.index and data.json files with the updated file and use the same data.readcount and eventalign.index files from the previous xpore dataprep output of HEK293T-WT-rep1 ?

Thank you for your help!

[yuukiiwa]

Hi @kwonej0617,

The data.json and data.index files are sufficient for running xpore diffmod.
Thanks!

Best wishes,
Yuk Kei

[kwonej0617]

Thank you for your reply @yuukiiwa !
In xPore paper, it was mentioned that nf-core/nanoseq was used to generate preprocessing data. I want to try to run nf-core from raw rast5. Could you please share your nextflow.confg and main.nf files?

Thank you so much for your help!

[yuukiiwa]

Hi @kwonej0617,

You will have to clone the nf-core/nanoseq pipeline, which includes both the nextflow config and main.nf files:

git clone https://github.com/nf-core/nanoseq.git

You can use the test config for RNA modification detection as a reference for the parameters and the the test data repository as a reference of the input directory structure.

Please join nf-core's slack channel if you run into problems when running nanoseq.

Thanks!

Best wishes,
Yuk Kei