2 DAYS OF LOOKING AT CODE, FINALLY FOUND THE BUG IN METHYLSEQ...I KNOW I SHOULD BE HAPPY ABOUT (& I AM) BUT I AM RANTING THAT IT TOOK ME SO LONG TO SPOT IT https://github.com/nf-core/methylseq/blob/54f823e102ef3d04077cc091a5ae435519f9923a/subworkflows/local/prepare_genome.nf#L51 (P.S: the bug was assigning wrong index ch for the aligner)

oh, so should have been

ch_bwameth_index =

rather than

ch_bismark_index =

yup

ooh amazing - is this the bug that has been repeatedly reported for like, a year now? 😅

looks like it

git blames it homer disappear

phew

looks like another bug, the fasta_index ch creation was nested in the bwameth scope https://github.com/nf-core/methylseq/blob/master/subworkflows/local/prepare_genome.nf#L64-L70

Is fasta index needed with Bismark? I had a feeling that was only used in the bwameth workflow.

ya right looks like its used only by bwameth. Thanks for confirming