How life sciences work: Findings of a year-long investigation (2019)
guzey.com
guzey.com
https://guzey.com/books/why-we-sleep/
The title is "Matthew Walker's "Why We Sleep" Is Riddled with Scientific and Factual Errors" and the contents is piece by piece refuting that book.
Checking google, it was discussed on HN actually: https://news.ycombinator.com/item?id=21546850
and in this sub-thread: https://news.ycombinator.com/item?id=21792342
Quite a few other potentially interesting links, even if only to check some things more deeply. Also, https://guzey.com/fiction/hntop1/ ("How I got to #1 spot on Hacker News and why you should never try doing the same")
My experience is that all labs fall somewhere between these two extremes, and within labs people are spread along this axis.
I felt a contradiction where the author states all ideas and execution comes from lab members, while quoting Boyden on the early days of expansion microscopy.
Even better is when the hosts are themselves working scientists.
Recently I got really into this podcast by Itai Yanai and Martin Lercher that tries to find how scientists navigate “day science” (eg formal papers and grants) vs “night science” (mucking about in the lab). Highly recommend if one wants to hear discussions by biological scientists talking about creativity.
Check out “Night Science”.
I imagine running a lab already takes 120% of ones time. I have no idea how Itai and Martin managed to do a podcast in addition.
This is entirely unsurprising to me. On top of this, reporters will see something interesting in a big name journal and report on it yet that usually adds a whole 'nother layer of incorrectness that the general public picks up as "fact"
Even if 10-50% of the studies I mentioned in the first section turn out to be wrong, the pace of correct and important results is still astounding."
1. These are all technological developments, not scientific developments. Even MRNA-vaccines leading edge stuff was being built when I was in grad school (20 years ago). So judging by what technical developments are coming out now, may be a lagging indicator on science.
2-1: Absolutely correct. Out of 10 years in science I only did 1 "officially on what we were funded to do". The other nine were on sneakily independent shit, or on fully independent (hard money) experiments.
2-2: No comment either way, have no experience with 'methods development' at least in the sense that he's talking about.
2-3: correct in the short term but wrong in the long term. If you are foolish enough to be a postdoc that puts everything into the science, you won't get promoted, so your track record of being a good scientist ends on the vine.
3. Correct in the short term but wrong in the long. I did exactly this but couldn't keep going because being a postdoc was terminal for me.
4. Correct. I tick off at least four of these bullets, including "unwelcome demographics".
5-12. All correct, without much else to add. Well said!
The biggest problem in sciences (biological) in the US is that we let the good quality scientists burn out and quit, and those who advance are mostly people who are playing the game (with a negative selection for those who are actually good scientists, because playing the game is so competitive at this point that it burns time effort and brainspace). This is a leadership defect. We don't give professors instructions on how to groom, including putting time, effort, and political capital into their grad students and postdocs to become lab leaders in their own right (if that is what they want)
One thing that the author does not (and cannot) address is deep knowledge, or thinking about things from first principles, a la feynman or musk. I think as we have more and more interdisciplinary scientists jumping in at the interdisciplinary level, we're already in an era where diletanttery (especially in trendy science mashups like biophysics) is high. This will further dilute expertise and make it hard to separate the wheat from the lemons. Once had a coworker grad student think you could drop a bacterial plasmid into a mammalian cell and get protein expression. I told him I would do it for him, but "I did not think that would work". He then confidently reported a positive observation (which, thank god, did not make it into any sort of publication). Dude is now a associate professor of genome sciences at university of washington.
what percentage of grant budgets "pad" for unofficial research, in your opinion?
this implies actual research could cost less and be allocated in smaller chunks if we tolerated higher risk and funded scientists more liberally.
decomposing the funding process into smaller "functions" would offer several benefits over the current model of one monolithic grant.
Once heard a bioinformatics (CS) PhD candidate say "natural selection would remove a region of the genome if it didn't have biological function". They graduated with a thesis having to do with genome evolution.
Over the extremely long run regions of the genome that have no biological function should disappear, no?
I think the only argument you can make is that 'eventually a neutral deletion would delete it' but I don't think we have a sense of what the timescale of that should be, and anyways, that isn't 'natural selection', unless you have some sort of argument that there is some intrinsic burden to maintaining DNA. If there is such a burden, it's very low, and probably doesn't affect the reproductive fitness of most organisms (esp. not higher organisms).
I think maybe for a really small organism, like on the order of Mycoplasma, or viruses, the carrying extra base pairs burden can be real, but this is not a general rule across biology.
even if they serve no purpose except to fill space, the added base pairs create "decoy" areas of DNA and decrease the likelihood of random mutations affecting vital functions.
If you have a sequence that's neutrally mutating with no corrective pressure it won't eventually disappear. At most it'll mutate into something unrecognizable [^1]
For example, your genome is littered with pseudogenes. These are mutated, ineffective shadows of what used to be a functioning gene. Nonetheless they're still close enough to what they used to be that you could reason about them in an evolutionary sense, eg: this one is also a pseudogene in other primates (i.e. it stopped mattering a long time ago), or this other one has a functioning homolog in non-primate vertebrates (i.e. it kept functioning in that lineage).
[^1] as a first order approximation. There are of course other structural events that can create much more change per event than your run of the mill single-point mutations that occur during every DNA replication.
I'm not a life scientist, but a physicist developing scientific equipment. My impression is that every branch of "science" has a technological side, to an extent that varies from one branch to another. Making and testing new therapies has been undertaken under the umbrella of life science for a long time. I think one reason is that the same "hands" that are good at doing the basic science work are also needed for technology work in those fields.
Another is that the underlying science isn't robust enough to expect a scientific development to be handed off to an engineering department without needing to learn more science along the way.
It's not like physics, where a mainstream engineer working on a technology is highly unlikely to discover gaps in physics theory.
I'm not really aware of any of those other than the gain of function research on bat coronaviruses by EcoHealth Alliance, which was only controversial beacause Fauci denied it in front of Congress and was technically not supposed to happen due to Obama's moratorium. What are the other revelations are you talking about?
The fact that this careful practical system still pumps out so much bs is very problematic. I think it's the same problem as other research areas, not the fault of the NIH system itself.
breakthrough ideas typically come from unconventional ideas, i.e., risky ideas or ideas likely to fail -- and thus waste taxpayer money.
yet the funding process is mostly about conforming to traditional assumptions, which naturally gates scientific breakthroughs.
the problem lies upstream. politicians don't allow the NIH to fund risky research and waste money.
how about NIH funding those gain-of-function coronavirus experiments (which included human testing) in Wuhan? And when DARPA refused to fund as too risky the additional human targeting coronavirus genetic manipulations in Wuhan, it was NIAID (led by Dr. Fauci - the top politician in life sciences today) which funded it.
Politicians expressly disallowed those experiments. Don't stir stuff up just to stir. It was a non-political agency exec (Fauci?) who apparently sidestepped the typical review process where they would have been killed, and funded them anyway
I also learned from that if you want to get funded by a study section, you volunteer for that section for several years and then you know exactly what to say to get funded.
Moreover, NIH funding, due to its central role in supporting biomedical research, is often necessary for getting a tenured PI position. Many high caliber R1 schools require multiple R01 NIH grants for tenure. Often private funding sources are less valuable to universities because they pay lower indirect rates (<10% vs >40%) on sponsored research. As institutions are so dependent on NIH funding, the NIH can exert influence far beyond the research it funds directly.
[0] https://health-policy-systems.biomedcentral.com/articles/10....
My experience is that this lies closer to 50%. This is the real problem with academic research and it’s only getting worse over time