Error! should be at least 1 contig

[james-e-barrett]

Hi,

I get the following error running

./modkit pileup \
~/Data/nanopore/mysample.pass.sorted.bam \
~/Data/nanopore/mysample.pileup.bed \
--ref ~/Data/hg38.fa \
--cpg
> calculated chunk size: 6, interval size 100000, processing 600000 positions concurrently
> filtering to only CpG motifs
> attempting to sample 10042 reads
> Using filter threshold 0.7207031 for C.
> Error! should be at least 1 contig

It runs fine without the --ref and --cpg arguments.

Any help appreciated thanks.

[ArtRand]

Hello @james-e-barrett,

Could you check that the reference you're using with modkit is the same as the one you used in alignments (i.e. the contigs should have the same names)?

[james-e-barrett]

Hi @ArtRand,

The reference used was indeed different (somebody else did the alignment). In my .bam file the chromosomes are labelled "1", "2", etc., but in the reference genome I used from UCSC (hg38.fa) they are labelled "chr1", "chr2", etc.

Changing the chromosome names in my .bam file fixed this. Using the same reference genome throughout would presumably avoid the problem also.

Thanks for your help.

[ArtRand]

Great, feel free to re-open this issue if you have any additional questions or encounter another related problem.