m6anet-run_inference, IndexError: single positional indexer is out-of-bounds

[huawen-poppy]

Dear developers,

Thanks for the nice tool! I tried to run the m6anet-run_inference, but there is an error saying that IndexError: single positional indexer is out-of-bounds. Could you please help me to solve this problem?

I used below code to run:

m6anet-run_inference --input_dir ./apo/dataprep --out_dir ./apo/inference --infer_mod_rate --n_processes 4

Please let me know if you need anything from my side. Thank you in advance for your help!

Best, Huawen

[chrishendra93]

hi @huawen-poppy, can you show me the first few lines of your eventalign.txt and data.info? I am assuming you're using the latest version of m6anet?

[huawen-poppy]

Thank you for getting back to me. Yes, I am using v1.1.1 m6anet. The header of the eventalign.txt looks like below:

I don't have a file named data.info. Could you please tell me where I can find it?

[chrishendra93]

hi @huawen-poppy , sorry I meant data.readcount and data.index

[huawen-poppy]

Hello @chrishendra93, thank you for your reply. Below are the first few lines for the data.readcount:

For the data.index file, it looks like:

[chrishendra93]

hi @huawen-poppy , the output looks fine, can you check if there are any samples with n_reads >= 20 in data.readcount?

[huawen-poppy]

Unfortunately, the maximum number of n_reads is 3 in data.readcount

[chrishendra93]

hi @huawen-poppy, I think you encountered the error because m6anet cannot find any positions to make inferences from. Currently m6Anet requires at least 20 reads from each position in order to make prediction

[huawen-poppy]

Ok, I see. I got this error by directly using all the original fast5 files. I tried to use the pass-quality fast5 files classified by Minknow tool to rerun the analysis. I can successfully run the process. It seems like using all original data directly would cover the signal.