By definition of complement,
Pr[not P | G and T] = 1 - Pr[P | G and T]
and by definition of conditional probability,
Pr[not P | G and T] = 1 - Pr[P and G and T] / Pr[G and T]
Pr[not P | G and T] = 1 - (16/100) / (33/100)
Pr[not P | G and T] = 1 - 16/33
Pr[not P | G and T] = 17/33