Without having read the PNAS paper or any of the cited papers, I could venture a guess that perhaps some papers were published on work still in the stage of getting the models to work. Of course the idea of such modeling should be to compare (calibrate?) to data on organic life, but they might not be that far just yet.

But of course, without any comparison to, or prediction of, organic life ever, the point of doing simulations is moot.

If the models truly never moves beyond the computational stage, then I find that appalling. Why on Earth would they not? (Okay, I know the answer, so don't bother.)

"Deep down the problem is that these guys aren't biologists at heart."

The problem is that these guys are not doing science. Period. Of course, the problem is not restricted to computational biologists. There are plenty of other flavors of biologists who don't do science. For example the entire field of genomics.

A more serious problem with modeling is that we don't have a proper framework to interpret the results. Even if you have a serious scientist who comes up with a model that's consistent with existing data and shows robustness. It may even make correct predictions. But is it the only such model? Maybe there are millions of models that explain the data equally well. How can you be sure? Exhaustive enumeration is impossible for all but simplest models, so we are stuck. This is a general problem, but especially severe in computational biology. Even for a problem as well-studied as sequence alignment, we have no good way of dealing with it. We simply retreat to the old standby of comparing our model against the random chance and try to impress other people with our p-value.

I work in a primarily computational biology group (called The Battelle Center for Mathematical Medicine) and I would say many of my colleagues would agree with the general idea that many papers would be substantially better and have a much greater impact if biologists were more involved in the work. I don't think a compsci should try to be a biologist any more than I want purchase an operating system designed by a PhD level biologist. I do think working as a team would be much better for both fields. There is however, a real problem with silos here. Compsci's are usually in compsci departments that have no biologists and have extremely different goals for their faculty. Biostats are in biostat departments. Stats are in stat departments. They teach classes for tenure. Writing a book isn't career suicide. Quick publications aren't necessarily how they want to spend their time. While there are computational biology centers, such as where I work and of course many of the large institutions have them too, there aren't nearly enough. There also aren't nearly enough biologists interesting in learning how to communicate/collaborate with computational people. I suspect in time more biologists will see the benefits of collaborating with computational people and the fields will slowly merge together in a more cooperative manner. Having good computational analysis of your real data can be helpful. I've published in good places and my colleague down the hall recently had a Cell paper - we both include lots of real data in our work. If you have any interest and don't mind learning how to communicate your ideas to someone outside of biology you should drop by a few local departments and see if a computational person sees something cool in your work. It may pay off better than some comments above would lead you to believe.

Its equally easy to complain about many of the small scale "careful" biological studies done where someone discovers that if you knock out component X, you find that protein Y appears to play a role in pathway Z. Who cares if that function actually matters in wt or that you only saw that in 1 out of the 10 westerns that were done, but 9 were discarded because they were clearly wrong. There is both good and bad science done whatever the methods used ( small scale, high-throughput, computational). These methods are useful under different contexts and have all made outstanding contributions to science.