Hi Diana,
The reason why its maximum is set to 1e-5 is because FUMA is not designed to annotate all SNPs in the genome so if there are lots of peaks and set to P=0.05 for example would cost a lots of computation time.
This won't be a problem in your case (since you have limited number of genes) but I cannot change this setting for specific person.
Alternatively, you can use fake P-value for FUMA input, for example you could multiply 1e-5 for P-value of SNPs with P<0.05.
I'm sorry for the inconvenience.
Best,
Kyoko