In engine assisted correspondence play it's never a good idea to play openings from engine books or human databases. Top rated correspondence players spend a ton of hours deeply researching openings and in order to be successful at that level they are limited to a very narrow choice of openings. And, once out of the opening, if one plays the engine's top choice every time one won't advance very far. You know the advice...blah, blah, blah. Boring, boring, boring!
As you also know, engines prune moves that they initially determine to be inferior, but sometimes pruned moves can contain a hidden resource. Here's how to discover some of those moves.
Use the engine to explore moves that look good to you. Unless you're rated north of 2500 they will probably not be very good, but not always. The other way is to look at moves the engine might have pruned. Here's how you do that.
Assuming your chess program allows it, you remove the engine's top choices and have it evaluate the remaining moves (except for obviously bad ones). This forces the engine to look at moves that it might have pruned. For example, in the position shown the top choices are 12.O-O-O, 12.f5 and 12.Bd3.
In this case, using Chess Assistant I went to the Infinite Analysis command and marked the moves I wanted the to force the engine to analyze:
This forces the engine to look at moves that it might have rejected out of hand and sometimes it will find something it initially missed.
Another trick is after the engine has had time to analyze the position step forward several moves in the best line and then start moving backwards looking for alternate moves the engine may have pruned.
These two tricks will sometimes help in finding hidden resources. It will require some time, but it can be interesting tinkering with the position.
I sometimes have opponents with ICCF titles and when I get a move that does not show up as one of either Stockfish's or Komodo's top three moves it signals that the position requires extra attention. I used to assume that they were using superior hardware or had let their engines run overnight or some such. After learning of the above two tricks I realized what had happened...my engines probably missed something in the time I let them analyze and digging deeper into that particular position was necessary. Happy mining for hidden gems!