I thought I could get by with that.
Sorry.
you examine something in a model system, and then it doesn't work out when you look at it in a human cell.
But that only means that your model didn't, well, model the reality accurately enough, didn't take all the important processes into account. That's why learning what to take into account and what to neglect is so important and difficult.
the point of research was to see if unexpected things happen and unless the computers involved are very clever, you can't tell it to look for something you never thought of looking for.
On top of what ita said, there are cases where, even if you set the rules yourself, you have such a big or complex system, that you can't foretell how the results will look like.
I'll give you an example. For my Master's, I worked on a simulation of traffic. Cars responding to each other. We had an equation to describe the accelaration of each car (wanting to stay at a certain time distance from the car ahead, not to have its velocity all that different from that of the car ahead - considerations like that). We put a few dozens of these simulation-cars on a simulation road, all of them having the same distance from each other and the same velocity. We let them ride for a while. Now, one would expect them to stay at those fixed distances and velocities, right? There's no change in the system, no reason for them to behave differently, right? However, they don't. Try this simulations with more than 30 cars, IIRC, and pretty soon you'll see jams appearing. We never put them in, but they were created anyway. They are also created IRL. I can explain why, if the post didn't make anybody run screaming looking for pictures of cows in tutus or petting chikens, but the bottom line is, you put the rules you thin govern the situation into a system, and you mostly have no idea what the results will be. On occasion, the only way for you to check those things is through computer simulations.
OK, shutting up now.