Previously, I wrote some comments about "Low Input, High Throughput, No Input". And I appreciate that Chen give many comment to my article, which inspired me a lot and made to think more about the problem of mechanism-based and data-driven research.

As I was doing the mechanism-based research, I guess I probably had some preconceptions about data-driven type of research. And I reconsidered about both two area for some time. The real situation is that in research there is not so many definite lines or boundaries between areas. Typically, the angles of research is usually debated from different lab. As for me, I am more practical for many things, whatever methods is OK as long as it could provide more and detail knowledge or critical principles explaining problems.

In both sides, there are many difficulties. For mechanism-based research, the not-easily-measurable parameters are usually obstructions. And being far away from detail mechanism-explaining makes data-driven part controversial. Thus debating is meaningless, finding a way out, no matter what method is applied, is more important. And I believe that there is potential power in combining these two types of research.

People talks so much about systems from the end of last century, but so far we don’t have any breakthrough in very large scale systems. I am not throwing cold water on this issue. I guess the big problem is that we need to contribute more intelligence for solving it rather than concepts bringing more issues.

Sometimes, I am quite frustrated by the reality that we can barely do any big steps. I feel there is a way out of this chaos and complexity, but I just don’t know it.

PS: In recent issue of Cell, there is an article about "GWAS makes it functional". I guess many friends of mine would be interested. Anyone mind if send me a copy? I can not access the database in KAUST (speaking of this, KAUST has so much money, why library database so sucks!!!)