Great 15-Year Project to Decipher Genes Stirs Opposition (1990) | Hacker News Reader

Great 15-Year Project to Decipher Genes Stirs Opposition (1990)

nytimes.com

14 pointsraldi16 comments

> They also doubt that the project can be completed in anything close to its original deadline and budget.

Spoiler: It was completed two years ahead of schedule and 10% under budget.

I suggest you view it in a different light. In the 1980s, as this article points out, the human genome project was funded by the DOE and worked on at the national labs. The NIH came in and earmarked money just for the project. This pushed it from being second class citizen to being on a pedestal.

Much of the doubt expressed was based on the experience of the previous 10 years of the project, and the changing budget situation that took money from other fields.

As a somewhat distant parallel, the HST, once fixed, produced great astronomy. How much other great science wasn't funded to put up the HST?

> Opponents doubt that the human genome project will quickly cure any diseases, either.

Spoiler: they were right.

> Much of the doubt expressed was based on the experience of the previous 10 years of the project

Are you sure? Here's an update from 1998: "The odds of success at this point are not overwhelming. At the end of this month, the project will be halfway through its planned 15-year course, yet only 3 percent of the genome has been completed."

http://www.nytimes.com/1998/03/10/science/the-struggle-to-de...

I'm sure that the doubts in 1990 were strongly affected by the HGP as carried out under the DOE days.

I'm sure the doubts in 1998 were based on different factors.

Internally, the HGP in 1998 considered themselves on track. Quoting from http://www.genome.gov/10001475 :

> The Human Genome Project has successfully completed all the major goals in its current 5-year plan, covering the period 1993-1998. A new plan, for 1998-2003, is presented, in which human DNA sequencing will be the major emphasis. An ambitious schedule has been set to complete the full sequence by the end of 2003, two years ahead of previous projections. In the course of completing the sequence, a "working draft" of the human sequence will be produced by the end of 2001.

How much other science isn't funded because we spend far too much money on other stuff? I'm sick of seeing engineers and sciences fight over a penny and demeaning the importance of each other's projects while we spend trillions on wars and other nonsense.

If you want more funding for other science projects fight to increase science budgets overall by taking money from stuff we don't need like yet another aircraft carrier. How many science projects weren't under because of just one of those?

Due to the long term nature of science benefits it's impossible to know the real importance of any one project anyway.

Cannot +1 hard enough. The F-35 will eventually cost $1 trillion.

> The F-35 will eventually cost $1 trillion.

Yeah, but for over half a century of procurement and use, and including all operating costs and inflation.

So, the equivalent of over half a century of the DARPA budget ($3 billion/yr) five times over, or two National Science Foundations ($7 billion/year), or half of another National Institutes of Health ($30 billion/year).

The NSF alone is responsible for over 20% of public nonmedical basic research in the United States so for the price of one boondoggle we can increase that total pool of funding by 50%. For over half a century.

Or we could nearly double NASA's budget and give it an economy of scale it hasn't had since the Apollo program to carry out even more basic research into astrophysics, climate change and metrology, geophysics, flight, material science, and countless other technologies that actually trickle down into the public market

Definitely hard to believe 15 years later. Especially with the possibility (however distant) of the benefits that personal genome sequencing will give to patients: ability to live a lifestyle that can lessen the likelihood of chronic diseases and drugs that are more effective and have less side effects since they are more or less tailored to a patient's DNA. Plus, the patient will take the drug for a shorter duration and at the minimum effective dose based on their needs.

I hate to make you feel old, but 1990 was 25 years ago.

> since they are more or less tailored to a patient's DNA

This is still pie-in-the-sky, and we don't know if that dream is much more achievable than the flying car.

Using an argument by analogy, we have sequenced a lot of disease genomes, but it's not easy to turn that information into new drugs.

I like ambitious projects. It's incredible how many of them actually succeed.

This reminds me of https://en.wikipedia.org/wiki/Human_Brain_Project as there is also lot of skepticism if it's doable. But maybe we just underestimate human ingenuity.

Do you have a tally of the ones that failed?

Fifth generation computing - https://en.wikipedia.org/wiki/Fifth_generation_computer

Strategic Defense Initiative - https://en.wikipedia.org/wiki/Strategic_Defense_Initiative

Expeditionary Combat Support System - http://spectrum.ieee.org/riskfactor/aerospace/military/us-ai...

The Field Data Collection Automation for the US census - http://calleam.com/wp-content/uploads/US_Census_FDCA_Case_St... (the failure added $3 billion to the cost of the census).

How do you score the Space Shuttle? It didn't come close to meeting its design goals, but throw enough money at something (as we did with SDI) does result in something.

But those are all engineering projects.. In a way, it's a lot easier to parallelize work on scientific project than engineering project. Competition from the industry actually helped Human Genome Project.

Anyway, my bet is that we don't actually need a very big supercomputer to emulate what brain does. We just don't yet know why it is so efficient, we don't know the trick(s). And I think, there is a possibility, as Human Brain Project will evolve, we will get better at figuring out what really needs to be simulated and what is just a fluff. And this will make possible to scale it up even on current hardware. Anyway, at least that's why I think it should be supported.

You said "ambitious projects". Now you're limiting those to scientific projects.

Again I ask, "Do you have a tally of the ones that failed?"

Would it include the Superconducting Super Collider? ($2 billion for a hole in the ground).

Actually, it's a trickier question than that. What does failure in an ambitious science project mean to you? Do you classify the two failed Orbiting Astronomical Observatory (of four) as an engineering failure, or a scientific one, or both?

For that matter, what is on your list of ambitious scientific projects, and how many of them have a success which is not primarily due to being a well-run engineering project? That is, the success of the Large Hadron Collider surely is due mostly to being a successful engineering project.

My experience is that the high-end science projects like what you are talking about (eg, the so-called 'Grand Challenge' projects), are constructed so they cannot fail. They lack a solid, measurable goal, and any advancement is counted as a success. This failsafe means they are more likely to get political backing.

I thought about your comments a bit and I think what's actually interesting in projects like HG or BB (or "Man on the Moon") is perhaps not scale as much as the fact that they are open ended. We don't know how the solution will look like. And that means that the solution may look really difficult or complicated today, but it's also possible that it won't be that difficult in the end.

So what I mean, I think, is not ambition in terms of resources, but rather ingenuity.

Reply on news.ycombinator.com