So the answer to my second question is ‘yes’?
It’s no for the pure modeling, yes or no for folding of the individual strands and depending on what level of accuracy you want and on whether you’re looking at pure folding, partial folding, folding during synthesis (which would be the real process)… Note that your example isn’t even a single-molecule protein, but a complex formed by four individual proteins, a tetramer.
Folding during synthesis is a big monster, you don’t even need to be looking at something as huge as hemo to have issues with that. For starters, we’d need to know the exact synthetic pathway.
That differs depending on the methods used - you can do very detailed methods on small systems, for larger systems you have to decrease the level of details and precision and make approximation. e.g. simulating proton dynamic in explicit water vs. approximations that do not take into account individual solvent molecules. Ab initio protein fold prediction methods normally do not simulate the natural folding pathway. So deepening of the scale of the system you analyse, you use different methods, often starting with corse-grained methods to get a rough solution, when adding more precise, but computationally expensive methods to refine the models.
Lots of challenges remain in modeling reactions. While we’ve gotten pretty good at predicting which reactions will occur, and how quickly (kinetics) and what products will be produced, this is mostly true for gas-phase reactions, where the reactants aren’t too big.
Some active areas of research:
Reactions in solution, especially in solvents which interact strongly with the reactants/ intermediates/transition states. We pretty much know how to model this (eg, include the water molecules explicitly, as several folks have pointed out), but in practice, this takes a huge amount of computer time, so finding approximate, but accurate methods, which don’t require explicit solvent molecules (or require fewer).
Reactions of complex molecules with multiple potential products. For example, what happens when you burn a biofuel? Many different molecules are found in biofuels (and the molecules vary depending on the source), and they have lots of different functional groups. The kinetic scheme for combustion of even one (relatively simple) biomolecule can easily include hundreds of species and tens of thousands of reactions. Automating the prediction of the kinetics and branching ratios of these reactions is a major endeavor.
Interactions which involve lots of competing, weak (intermolecular) forces. Solvation, drug-protein binding, protein-protein interactions.Today’s In the Pipeline blog (by Derek Lowe) discusses some of the challenges remaining in modeling binding of potential drug molecules to targets (proteins, DNA).
That’s just a start…we’re nowhere near solving chemistry.
And here is an announcement of a new way to synthesize diamonds.
The Journal of the American Chemical Society is a higher-end American general chemistry journal that is published weekly. The table of contents should give OP examples of the sort of chemistry that people are still studying. Let me know if you have questions about any of the articles.
I’m not so good with the modeling. It is more or less useful depending on what question you are trying to answer. What, roughly, is the reaction barrier for this catalytic reaction? That is often not so bad. Which catalyst is more enantioselective? Useless (not that that stops people from trying.)
Philosophers stone is still a mystery, anyone finds out let me know.
I am a Chemical engineer and have spent a fair amount of time in my profession modeling reactors and reactions.
Even simple reaction chemistry is not well understood or modeled, otherwise people would not be paying $5 for a MMBTU of Methane and $18 for a MMBTU of gasoline when both are just hydrocarbons (and the former is abundant in the US)
Even for homogeneous reactions - the understanding is fairly limited. Take the simple reaction of Methane burning in air. To get an approximate flame pattern prediction, you need to model 325 elementary chemical reactions and around 53 species (a lot of public tax dollars went into the development of these). Even then, you cannot predict flame patterns reliably or under higher pressure - etc.