I think problem with most people esp when we are in realm of behaviour is what Sapolsky might call categorical thinking. We tend to try and explain this as completely dependent on nurture or genes, or neurotransmitter, or hormones or evolutionary history. But a simple behavioural act, say something as simple as a smile to a certain stimulus is all of it at once.

Are there social cues that train us that we should smile at a certain context A but not at context B? Yes

Are there neurons which fire and release neurotransmitters to make us smile for context A and inhibit it for context B? Yes.

There would also be hormones that we're secreted when something made us happy and something did not. There would be hormones that would predict our responsiveness to happiness.

Are there genes which if present can make us more sensitive to judging when to smile and when to not? I mean these could be genes involved in development of brain regions involved in judging social cues. There will be genes involved in learning social cues and responding to them. And that one simple simple in right context would be cumulative effective of all these genes?

But is having and missing certain genes tells the whole story? Well, genes have be expressed. The quantity of expression of various genes are under influence epigenetic control. A certain exposure to stress during prenatal and neonatal life may influence, say how corticosteroid receptor genes will be expressed. This may predict how one may repo and to stress for all their life. Whether they will be resilient, alert or anxious? This will eventually effect how they will respond to social cues. How shy they feel in smiling how how confidently they smile. This may also program to to smile in wrong context.

Is there an evolutionary reason why we smiled in certain context and why not in other? Perhaps judging social cues right might have improved fitness. Smiling right meant more friends or more mates, who knows.

I don't know why would gender identity, attachment, depression are debatable for biological origin. If it is happening in a biological system, there has to be associated biology.

It's like asking if an AI code written in a computer that made computer better at finding a certain pattern in data a function of computer hardware or it's programming post manufacture. Well, of course programming played a role. But remember not all hardware configurations would be equally efficient or capable for similar program. At the end of the day it's hardware that executed the program. For biology the hardware is even a bit tunable. Moreover, there would be some explanation for why a hardware compatible to efficiently execute the program came into being.