Should we include 0's when calculating q values?

Question

[solarchan7]

Basically, I have two sequences that the control groups are all 0's but the test groups have some values (none zeros); similarly, I have another 3 sequences that the test groups are all 0's but the control groups have some values (none zeros). So, I am wondering when calculating the q_values (there are ~50 sequences in total), should I include these rows, or maybe some of them?

Thanks in advance!

[e.rempel]

Please, provide more context.

[solarchan7]

my table (example) is like this: there are rows with test == all 0's / control == all 0's / normal should i include all of them when doing the q_value? table: https://ibb.co/bvZ1XFF