AgeVivo, on 11 September 2011 - 10:16 PM, said:
Would you think that RNA expression at various points in time (on specific tissues or on a mixture of the whole body, to be defined) and for the 3 frog species (Rana sylvatica (the cryopreserving&thawing frog), the already-sequenced frog, a closer cousin to Rana sylvatica but that can not cryopresenver&thaw well) is likely to bring smthg useful in the field?
What type for useful result could we perhaps get? Is ths feasible with $6000? Who would lead by whom in what lab/environment?
I don't know how much value the already-sequenced frog is. It diverged from Rana sylvatica's ancestors in the Cretaceous, when dinosaurs walked the land. (Frogs have been around for a long, long time.)
The purpose of sequencing a closely-related species is to find genes that R. s. has that the related species doesn't. An RNA exp
ression time-series for R. s. and a related species would, similarly, let us see how genes are expressed differently between the two species during freezing and thawing. This would probably be even better than sequencing, and possibly cheaper. Ideally, we'd like to measure protein levels, but we can't easily look for proteins without knowing their sequence first.
I don't believe it will be a simple matter of saying, "We need these three proteins to survive freezing". There's a complex regulatory program involved in freezing. This means many of the important differences will be not in the genes themselves, but in their regulatory sites, many of which can't be found with RNAseq.
On the bright side, if we take enough measurements to measure the time-progression of the different proteins involved, we might be able to reproduce this time progression in human subjects (after testing on others, of course) via intravenous feed, without having to solve how the regulatory network works. This depends on whether the freezing response is more or less the same across all tissues (good, easy), or whether it has to be managed individually for different cells (harder, requires gene therapy). And it also depends on whether you need to keep changing concentrations within cells after the circulatory system has frozen up; if so, you need gene therapy to make the cells do that themselves.
Either way, I expect it will be easier to modify humans to be freezable, than to modify our existing freezing techniques to work on wild-type humans.
I wouldn't start any new eukaryote sequencing project now. There are cheap new DNA sequencing technologies coming out this year that might change the cost a lot.
I also anticipate lower computer analysis costs, because people are going to have to stop their wasteful "BLAST every gene against every other gene ever sequenced" approach to figuring out what genes they've found. Currently, computer time to analyze a genome costs roughly as much as the sequencing cost. That's because it takes several CPU-years to analyze a genome, and you also have to pay the salaries of IT professionals maintaining a large, complex, distributed computing infrastructure, with expensive distributed storage, which seems to not have economies of scale - I think it's actually more costly per CPU cycle than a desktop. I believe you can do a thorough analysis on a single CPU at less than 1/100th of the cost; and people will have to figure this out sometime in the next 5 years, because the BLAST databases are growing much faster than disk storage or CPU speed.
Since we're on a budget, better to wait one or two years and re-evaluate.
As to who will do it - We do all these things in-house; but JCVI doesn't do sequencing for hire. I don't know why, but we don't. We work from grants. But if you proposed this as joint preliminary research to lead to bigger contracts, we might kick in a few thousand dollars, do the sequencing, and pay for the computer time, if we thought the results would lead to bigger NIH contracts in the future. For instance, the frog's response to prevent the glycogen it uses as antifreeze from killing it could be relevant to diabetes.