This is where it gets fun - the answer to those questions is entirely up to you! Your source sound can be a synthesizer, a vocal, a sound you ripped off youtube, a recording of a dog barking, or anything else you can think up.
From there, you can do lot of different things:
You could try applying layers of distortion, filtering, and limiting to get the more noisia style neurofunk basses. You could pitch shift and/or timestretch the sample, chopping it up and bouncing that and doing it over and over again. You could put it in a sampler, play random notes, radical EQ changes, compress it, then resample that and try it again. You could apply a lot of reverb or delay then distort and/or filter the resulting sound.
What's important is that you experiment with a lot of different methods, so you can learn how they affect your sound differently.