AN field in the phase_common phased data is different from the AN in unphased data

[lessdata]

I have an unusual observation regarding the AN field in the output file of phase_common. It appears that the AN field in the output file is calculated as the sum of AN (target) and AN (reference), which differs from the AN (target) value present in the input vcf file (target).

While I understand that AN may not be crucial for downstream analysis and can be recalculated if needed, I am curious to know if this discrepancy is intentional or if there might be something wrong on my side.

Thank you.