Questions about the m6anet inference results

[mumdark]

Dear Developer,

Thank you for developing such an great tool !

I have a question when running m6anet inference.

Why are there only a few positions, specifically 16, retained in the output? In my data, there are far more positions with more than 20 reads.

---

Here are the read counts for my input data(a total of 1,117 positions with more than 20 reads):

Here are my results:

Here's my code:

m6anet inference --input_dir m6anet_datapre/ --out_dir m6anet_inference/ --n_processes 4 --num_iterations 1000

Thank you for your response! Much appreciated.

mumdark

[mumdark]

It appears there was an issue when running m6anet dataprep for data preparation, which might have resulted in the unexpected final output.

How to solve this problem? Thanks.

[JSluo888]

I am having the same issue

[soniacruciani]

Hello,
I also am encountering this problem. I checked my .json file and it is fine, I have 14 positions which I know should be picked up (and were on a different fast5 file of the same exact experiment), but in my output I only have 13.
I have ~2k reads coverage per position, and ~800 reads at least for each position in dataprep output.

Any idea of what might be happening?

Thanks!