A new NIH RFA:
PsychENCODE: Identification and Characterization of Non-coding Functional Elements in the Brain, and their Role in the Development of Mental Disorders (R01)
The Encyclopedia of DNA Elements (ENCODE) project, by systematically cataloging transcribed regions, transcription factor binding sites, and chromatin structure, has recently found that a larger fraction of the human genome may be functional than was previously appreciated. However, our understanding of the role of these functional genomic elements in neurodevelopment and mental disorders is at an early stage. This funding opportunity will support studies that identify non-coding functional genomic elements and elucidate their role in the etiology of mental disorders.
Suddenly, the ENOCDE model is now the way to do science. It’s hard to disagree with Dan Graur on what the consequences are:
The big difference between the Human Genome Project and ENCODE is that the former produced data that are undisputedly true and useful (a C is a C is a C is a C). The ENCODE project is a collection of zillions of experiments, each employing a certain cell line, a certain set of temperatures, a certain pH, and many other particular experimental conditions. The results are not as clear as the sequence CTTGACTTAT. So, while all the world is interested in human genome sequences, and the sequences are an objective set of data (albeit not an errorless one), the ENCODE data may be useful for some and useless for others. I, for one, do not regard HeLa cells as human. I don’t think we can learn anything “human” from these cells or any other malignant cell lines…
The funding of ENCODE has cost many small-science scientists their scientific livelihood. Thus, many people outside ENCODE may well use the ENCODE data in the future simply because they have no money to do the experiments they wanted to do in the first place.
American science has been great for many years because it allowed small science and individual scientists to flourish. Now, it allows a few people who have learned how to con the system to flourish (usually if they manage to gather enough technicians to do their bidding).
Not only are the experiments being done in meaningless cell lines, the results are often incredibly sloppy. Since when has a correlation of 0.36 between experimental replicates become acceptable?
At some point, some people decided that doing ENCODE was a good idea. Now that the data is available, I will use it and derive some benefit from it. It is not clear to me, however, that this has been a net benefit for biomedical science. And seeing the ENCODE model set in as permanent, long-term model for how to do science is incredibly infuriating and disappointing for someone working out of a small lab, trying to build a career in what is becoming an increasingly dysfunctional field, in spite of the recent spectacular advances in technology.