I would like to estimate heritability and genetic correlation from my data using LD score regression. But, my data (summary statistics) has some SNPs with extremely low P values like P-value=5.0x10^-700.

If such SNPs are included, munge_sumstats.py outputs an error. To deal with this, should I apply munge_sumstats.py after excluding the SNPs from my data? Or should I just round the P-value to some higher value (e.g. P=1x10^-100)? I would appreciate your advice.