Couple of pointers:
First, as many have already pointed out, it just takes practice
Second, a simple but effective training regime is to start with simple sentences, and gradually work your way to more complex sentences
Third, do not see this as a one-way street from logic symbols to English. Rather, you should learn the two-way street between them. That is, you should learn how to translate English into logic at the same time as you learn to translate logic back into English; it is only yhen that your brain really learns to see the connections.
Fourth, keep a close eye to the grammatical structure of the sentences: see where the parentheses are, so you see the scope of the quantifiers, and the order of the operators. The translation obviously depends on exactly those things.
Fifth, many predicate logic sentences that you will encounter follow the basic structure of the four Aristotelean sentences, exactly because so many English sentences are of this format:
'All $P$'s are $Q$'s': $\forall x (P(x) \to Q(x))$
'Some $P$'s are $Q$'s': $\exists x (P(x) \land Q(x))$
'No $P$'s are $Q$'s': $\forall x(P(x) \to \neg Q(x))$ (which is the same as 'All $P$'s are not $Q$'s')
'Some $P$'s are not $Q$'s: $\exists x (P(x) \land \neg Q(x))$
Commit these four patterns to memory, and you'll find they pop up all the time in predicate logic sentences.