Resume function is re-doing alignments instead of resuming at methylation calls

Hi,

I am fairly new to Nextflow/nf-core, but nf-core/methylseq was up and running smoothly until halfway through the methylation extraction it crashed my whole server due to insufficient resource limits. I fixed this problem by being more strict with my custom config.
After I tried to re-run the program with -resume. However, it is attempting to rerun all alignment steps even though the first run had completed all of those and progressed halfway through methylation extraction. This is a major issue because the alignment step took about 2 weeks to complete, and I do not have enough time or storage to re-run the alignment.
From what I can tell, changing the config invalidated the Nextflow cache, and because I used the default settings (which do not retain alignment files), the pipeline now treats those outputs as missing and is rerunning alignments. I haven't been able to find a way to tell methylseq to skip alignments and start from the existing deduplicated bam files, and it doesn't seem possible to provide bam files as input instead of fastqs.
At this point I'm a bit stuck and was hoping someone here might have encountered this or have more experience to know how to workaround this.

Thank you so much

Start by reviewing how methylseq handles Nextflow -resume after a configuration change and how it represents alignment and methylation-extraction outputs. Compare the existing deduplicated BAMs with the pipeline's FASTQ-based entry path; done means identifying a supported way to resume at methylation calls without rerunning completed alignments, or documenting that no such path exists.