Zero reads in index error

[setshabaTaukobong]

Good day,

I am trying to detect base modifications but I keep getting a specific error.
I basecalled using the Promethion instrument and got multiple bam files.

I then merged these bam files using the following command:
samtools merge merged.bam *.bam

I then proceeded to sort and index the final file.
From there I used the following command:
modkit pileup Final.bam pileup.bed --log-filepath pileup.log
I however get this error:
calculated chunk size: 6, interval size 100000, processing 600000 positions concurrently
attempting to sample 10042 reads
Error! zero reads found in bam index

Could I please get some assistance on how to solve this?
Thank you

[ArtRand]

Hello @setshabaTaukobong,

Did you run alignment concurrently with basecalling? To run modkit pileup you need to have reads aligned to a reference. You can inspect the base modifications on a read level with modkit extract and modkit summary.

[setshabaTaukobong]

Hi @ArtRand
thank you so much for getting back to me.
So just so understand, I would have to provide the sequencing machine with a reference file during basecalling?

[ArtRand]

Hello @setshabaTaukobong,

You can perform alignment during basecalling by providing a reference, or if you've already basecalled the reads I would align them to the reference using dorado aligner. I would not re-basecall the reads if you don't have to - just align them. Unless, of course, you want to use a different basecalling or modified base model.

[ArtRand]

@setshabaTaukobong any luck?