@dmacguigan pointed out in gitter that in igv alignments of the consens.bam file were off by one from the reference sequence. After a while I figured out that in clustmap.build_clusters_from_cigars the call to bedtools returns 0 based positions, but then we build clusters and eventually construct sam files (1-based) using these positions without accounting for this. I adjusted it so that now positions in the clustS files are immediately converted to 1-based. Everything downstream is still correct, I think.
@dmacguigan pointed out in gitter that in igv alignments of the consens.bam file were off by one from the reference sequence. After a while I figured out that in clustmap.build_clusters_from_cigars the call to bedtools returns 0 based positions, but then we build clusters and eventually construct sam files (1-based) using these positions without accounting for this. I adjusted it so that now positions in the clustS files are immediately converted to 1-based. Everything downstream is still correct, I think.