Significant colocalisation across two GWAS SumStats with opposite "effect sizes"

Question

[Pine]

Hello Biostar Community,

If I run a colocalisation test for a locus across two GWAS SumStats (that are correctly harmonized) and I see evidence for colocalisation (Ex. high PP.H4 in coloc); but effect sizes are in the opposite directions, would it be reasonable to interpret this as an evidence for the locus having significant yet opposite effect across the two traits (or should I interpret this as evidence for lack of colocalisation)?

Thanks in advance!