This answer is from common knowledge rather than specific infield expertise. The question could be more expertly answered by other members.

If you really want to pursue this type of calculation its possible that something like `INPS-MD` *de novo* -> protein structure stability from variants will provide an answer. So maybe compare the outputs with `AlphaFold`? 

Intuitively I don't think its cool, just my opinion. In my opinion an experimentally derived structure is required as the basis for your calculation, where you want to consider the impact of an amino acid mutation.

Its important to separate *de novo* prediction from variant prediction. Alphafold is *de novo* prediction. What you are looking at is structure stability from a variant (see Summary). Thats my core answer, these different calculations. To look to explain this more clearly ... 

Good variant stability calculations are DUET (2014) and [INPS-MD][1] (2016), amongst others. Just to reiterate a good starting point is an experimentally derived protein structure.

The webserves are here, 

* [http://structure.bioc.cam.ac.uk/duet][2]
* [https://inpsmd.biocomp.unibo.it/inpsSuite][3]

I personally don't know `Alphafold` but understand its predecessor's which accommodate convolutions by leveraging deep-learning. The basis of the predecessor's calculations firstly are *de novo* prediction and secondly can miss the impact of variants if they are not directly involved in the core-algorithm. Its about identifying which amino acids are interacting considering all available homologous amino acid within a database. The calculation can be predict local interactions with extreme accuracy, but its a prediction. It will also depend on whether there are a single amino acid change or a two- or more amino acid changes (they could compensate each other particularly for a natural variant). 

In contrast DUET and INPS-MD amongst a lot of other similar programs consider the property shift of an individual amino acid in relation to its immediate local stereo-chemical environment and measured by a shift of free-energy. 

Essentially, these are two sets of very different calculations.

**Summary** The problem I intuitively see is if *de novo* prediction in any form is then used, okay they are very good, but they're just predictions. Variant structural prediction are also just predictions  which are very widely respected. Stacking *de novo* prediction onto predicting protein structure stability resulting from variants has scope for error, especially if there was some hidden non-linear effect.

There could be additional issues because the *de novo* calculations immediately prior to `Alphafold` doesn't predict the precise local distances.

-------

`AlphaFold` would is simply respond 
>'its not been trained'

They'd want a wet-lab experimental dataset with wild-type vs. variants, if these exist with sufficient abundance to prevent overtraining. It would be a very skewed dataset.

  [1]: https://academic.oup.com/bioinformatics/article/32/16/2542/1743481?login=false
  [2]: http://structure.bioc.cam.ac.uk/duet
  [3]: https://inpsmd.biocomp.unibo.it/inpsSuite