To really nitpick, first you'd need to factor in the varying digestions
for each type of protein (which can very from a low 70% for vegetable
sources up to about 97% for meat and protein powders) to factor in the
amount of protein that's being lost in your feces. So now we need to
find
Actual protein digestion = Sum (P1*D1 + P2*D2....Pn*Dn)
where Px is the protein in question and Dx is the digestibility constant
for the question) and where n is the total number of different proteins
being consumed.
Now, the Atwater constant for protein varies from roughly 3.9 to 4.2
cal/gram (4 cal/gram is an average) so you'd really need to find the
exact constant for each dietary protein you're eating, and multiply that
by the gram amounts of protein of each.
So we now have:
Net caloric content = Sum (P1*D1*A1 + P2*D2*A2...Pn*Dn*An)
where Px is the protein in question, Dx is the digestibility of that
protein, and Ax is the Atwater constant for the question) and where n is
the total number of different proteins being consumed.
Of coures, that only tells you the caloric content (as measured by a
bomb calorimeter) for each protein, while factoring in digestion. This
doesn't tell you how much energy is actually derived by the body (which
is NOT a bomb calorimeter).
To get an accurate number, you'd need to take into account the fact that
the energy yield from protein is far lower than from other nutrients
(i.e. carbs and fat) so you'd really have to take the number of carbons
in each protein and track them through the various metabolic reactions,
counting the carbons yield all the way, to find out how much of the
actual absorbed protein is yielding energy.
Or just ignore the bored geek and use 4 cal/gram.
J'aime assez la dernière phrase
Le problème c'est qu'on ne peut pas comparer l'énergie récupérée dans un calorimètre et celle utilisée par notre corps... comme toujours.
mais je sens que je vais le faire 
