All the 20 amino acids are composed of periodic element motifs that have determinable sets of reactions to the proximity of other elements. Thinking dynamically, if these elements are in turn made of subatomic particles & then quantum particles with characteristic profiles, then the magnitude of reaction freedom becomes immense. That's not taking into account environmental factors like pressure or temperature but only what things are and their relationship to each other, including properties such as electronegativity, hydrophobicity, polarity... which all generate the biochemical behaviours of proteins.
Think hydrophobic amino acids would collapse internally into the protein or allow it become membrane bound while polar amino acids would do the opposite unless of opposing charge to the proximal elements, and cysteines form intramolecular disulphide bonds. As von Neumann said about mathematics, I'd say for the chemical behaviour of amino acids; You just get used to them with time, exposure and curiosity.
As Anderssen said above, even though the fundamental interactions are (more or less) known, their effective computation is extremely complicated.
However, a new way of looking at this problem is through the evolutionary theory : A given protein appears (with mutations) in a large number of species. Sequences which are conserved, and specially the correlation between these sequences, contain huge amount of information about the influence of each amino acid on the protein. See for example the following paper
Cell, Volume 138, Issue 4, 774-786, 21 August 2009
( a pdf version can be found by googling "protein sectors" ).
Ok I am trying to simply explain it. Well that depends... If you e.g. look at an enzyme it will highly matter if you change an amino acid exactly where the enzyme reacts. If you change it somewhere else it might obviously matter much less (so you can only change a certain number of them otherwise the protein will not work, I think this is what you are referring to). The other thing that matters is what you exchange an amino acid with. Some are more similar than others. If you exchange two very similar ones it will make less difference than changing two totally different ones.