One example is NM_000059.4:c.5del, with start==32316463 and end==32316464. This is a single-base deletion, and the start and end should be the same. See https://www.ncbi.nlm.nih.gov/clinvar/variation/51985/?oq=NM_000059.4:c.5del&m=NM_000059.4(BRCA2):c.5del%20(p.Pro2fs). The canonical SPDI is NC_000013.11:32316463:CC:C.
In ClinVar.txt, the "genomic coordinate" was estimated as chr13:32316463:GC>G.
In aggregate_across_columns.py, start is set to pos, and end is set to start + len(ref) - 1.