I'd probably stick with what you know. If you were doing Ruby, I would've heavily recommended Ruby on Rails -- although a ton of people talk down on it being "opinionated", I think those opinions are within reason -- having been vetted by a LARGE open-source community of very experienced developers. I think it puts a person starting out on solid ground from which to start crafting their own viewpoints about what seemed difficult and what seemed hard and why it differs in other frameworks or paradigms. The NodeJS community unfortunately is immensely fragmented and difficult to wade through when it comes to writing software on the back-end or full-stack.

I always urge first principles and literally search for "what is X?" whenever you don't know something and follow it up with, "why is X done this way?" or "why does X exist?". There's no way to go down every rabbit hole, but it this allows you to start crafting branches of knowledge for future leaves to hang on.

The thing that I implicitly use everyday is design principles and general algorithmic thinking -- mostly running time analysis. When I say design principles, I'm kind of alluding to "clean code". This is a popular repository for JS and non-JS people alike: github.com/ryanmcdermott/clean-cod.... I would star it and use it as a reference. It does put into perspective some things -- like why some things are called what they are. The back-end of things is "everything". One of the most invaluable books I've read across my career in software thus far is, "Designing Data Intensive Applications" by Martin Kleppmann. I think it's a good starting point for setting up branches even if upon first read, half the things are questions marks do you. It's best to have at least covered some basic algorithms and data structures, though