Sometimes gibberish gets published.

[chadn737]

It seems that these two physicists just now learned about genetics and decided that they would overturn an entire field with gibberish.

 

The sequencing of the human genome raises two intriguing questions: why has the prediction of the inheritance of common diseases from the presence of abnormal alleles proved so unrewarding in most cases and how can some 25 000 genes generate such a rich complexity evident in the human phenotype? It is proposed that light can be shed on these questions by viewing evolution and organisms as natural processes contingent on the second law of thermodynamics, equivalent to the principle of least action in its original form. Consequently, natural selection acts on variation in any mechanism that consumes energy from the environment rather than on genetic variation. According to this tenet cellular phenotype, represented by a minimum free energy attractor state comprising active gene products, has a causal role in giving rise, by a self-similar process of cell-to-cell interaction, to morphology and functionality in organisms, which, in turn, by a self-similar process entailing Darwin's proportional numbers are influencing their ecosystems. Thus, genes are merely a means of specifying polypeptides: those that serve free energy consumption in a given surroundings contribute to cellular phenotype as determined by the phenotype. In such natural processes, everything depends on everything else, and phenotypes are emergent properties of their systems.

 

Yes genes are means of specifying polypeptides....in addition to how and when they are expressed. What they seem to ignore is that differences in polypeptides, which underly phenotype, are encoded in the DNA and thus inherited by DNA. That is what Natural Selection acts upon. This is a perfect example of the nonsense that can result when people jump into a field new to them and think that they understand it better.

 

The guest blog post makes even less sense.

[Endy0816]

Yeah, I've seen some questionable papers in even the more reputable journals.

[Strange]

Reminds me of this: http://xkcd.com/793/

[CharonY]

That xkcd popped into my mind immediately. It is not that uncommon, unfortunately. But I also had very positive collaborations in which all sides at least made a token effort in trying to understand each other. Obviously having access to copious amounts of beer and coffee helped.

[timo]

Reminds me of this: http://xkcd.com/793/

Incidently, I spent half of the day looking through data of how well my simplified model (of a non-physics system) describes reality. At some point I realized that my excuses why data don't match my model became a bit ridiculous, so I ended up throwing them all into the trash bin (the excuses, not the data) and going with "doesn't work, but we don't have a better alternative so we stick to the model" . Oh, the joys of contract-based research where every project has to be a great success by definition.

[CharonY]

Oh, but in academia everything is a success, too. Problem is only that there are not enough journals such as Applied "why the heck dioesn't it work? It did last time?"; Journal of the Royal Society of "the new student did what? Why is he/she still alive?" and Proceedings of "which hypothesis can we make up to fit the data?"

[ajb]

Oh, but in academia everything is a success, too. Problem is only that there are not enough journals such as Applied "why the heck dioesn't it work? It did last time?"; Journal of the Royal Society of "the new student did what? Why is he/she still alive?" and Proceedings of "which hypothesis can we make up to fit the data?"

This is for sure a problem, especially when you are early in your career. Talking to experts in your field can help with this, but no single person knows everything and I am sure some old ground get trodden on over and over again.

 

The only exception to this I can think of are "no-go-theorems" in mathematics and physics where there is a strong result as to why something will not work. More mundane failures don't get published.

 

Related to this is rediscovering work, again when you start out this can be common.

[swansont]

Reminds me of this: http://xkcd.com/793/

 

The thing is, as I have realized, is that some physicists doing this are just approaching the problem from a different perspective. We're used to not being able to look inside the black box so we make models, while the other scientists are actually trying to work out the mechanisms, because they actually can.

 

But turn this around. Are the biologists analyzing the problem by analyzing the 2nd law, or principle of least action, or any other physics principles (which must hold) to see if that affect the outcome? It may not add anything to the solution, but you don't a priori know if that's the case.

 

I remember a biology result in which they were puzzled about some behavior (why people walk in circles when there's no reference point) where some basic physics and noise analysis shows that the null hypothesis that we would walk in a straight line is incorrect. So complaining about physicists poking their nose into biology is one thing, but if that means that biologists are ignoring physics then they're in the wrong.

[chadn737]

 

The thing is, as I have realized, is that some physicists doing this are just approaching the problem from a different perspective. We're used to not being able to look inside the black box so we make models, while the other scientists are actually trying to work out the mechanisms, because they actually can.

 

But turn this around. Are the biologists analyzing the problem by analyzing the 2nd law, or principle of least action, or any other physics principles (which must hold) to see if that affect the outcome? It may not add anything to the solution, but you don't a priori know if that's the case.

 

I remember a biology result in which they were puzzled about some behavior (why people walk in circles when there's no reference point) where some basic physics and noise analysis shows that the null hypothesis that we would walk in a straight line is incorrect. So complaining about physicists poking their nose into biology is one thing, but if that means that biologists are ignoring physics then they're in the wrong.

 

I'm not saying to ignore physics or that physicists have nothing to contribute. I would argue the opposite. I also value taking a different look at things. What I do not value are grandiose claims that ignore a vast body of research and also seem to miss the basic point.

 

The essential gist of this paper is that inheritance and thus evolution has little to do with genetics, but the passage of physical states. What it is, is essentially a neo-Lamarkian "epigenetic" view. They cite only a handful of examples to support their case, but ignore 100 years of extensive genetic research and dismiss it quite out of hand. Part of their argument is that variation comes from protein folding....therefore genetics is not really relevant. What they ignore is that protein folding is determined in no small part by the polypeptide sequence and its variants which itself is determined by the DNA.

[CaptainPanic]

There is a difference between:

 

1. You can do good science, but achieve disappointing results (e.g. process turns out to be not economic), and then pimp the results with some unreasonable assumptions, so you don't endanger the next round of funding.

 

And:

 

2. You can also just do bad science, and still publish.

 

I notice that nearly 100% of PhD students will graduate, most of them also within the 3 or 4-year span that they are paid by the university. And both the PhD student as well as the university have an incentive to publish anything: the PhD student needs a minimum of X publications within the 3 or 4-year span of the contract, and the university has a 'Key performance indicator' of the number of publications that they produce per year.

 

The core of the problem is that funding bodies demand that money generates great science, but are incapable of actually going into the details - and they don't trust the scientists to do good science by themselves. So, they've set some performance indicators. And the scientists dutifully oblige and just publish something.

[Arete]

I notice that nearly 100% of PhD students will graduate, most of them also within the 3 or 4-year span that they are paid by the university.

 

 

My anecdotal experience in the Australian system may not be representative, but I saw a number of students either drop out or downgrade to a Masters. My experience in the US differs in that I've seen students pull out early at the prelims stage, or finish (even if they stretch it out the eight years). The difference being that in Australia, you have a hard deadline on your stipend, where as in the US (in my limited experience) the duration for which you can be paid as a grad student is more squishy.

 

That said, there's still a huge range in the quality of work that's considered acceptable to make up a PhD dissertation, even among students with different committees at the same university, and plenty of people out there with a bit of paper and a title who don't necessarily know what they're talking about (and people without them who do).

[CharonY]

 

The thing is, as I have realized, is that some physicists doing this are just approaching the problem from a different perspective. We're used to not being able to look inside the black box so we make models, while the other scientists are actually trying to work out the mechanisms, because they actually can.

 

But turn this around. Are the biologists analyzing the problem by analyzing the 2nd law, or principle of least action, or any other physics principles (which must hold) to see if that affect the outcome? It may not add anything to the solution, but you don't a priori know if that's the case.

 

I remember a biology result in which they were puzzled about some behavior (why people walk in circles when there's no reference point) where some basic physics and noise analysis shows that the null hypothesis that we would walk in a straight line is incorrect. So complaining about physicists poking their nose into biology is one thing, but if that means that biologists are ignoring physics then they're in the wrong.

 

Absolutely. However, if you bring your expertise to a complex problem it is worthwhile to actually try to establish the foundations first. The reason why biology exists at all is that we are not able to rebuild the system based on first principles. As such, biology has developed methodologies that are less precise, but can at least tackle the questions to some extent. The point where progress happens is not where biology does fundamental physics and vice versa but where the problem is dissected into components that can be optimally addressed by the different methodologie (and ideally assembled into a coherent body of knowledge).

 

What the authors in the OP did where recasting a biological problem completely in different terms that actually do add little to its solution. In a way they are revising a straw man. What they (correctly) challenge is a deterministic flow from gene to phenotype. At the same time, it ignores another point typical to biology is that it does high-level simplifications with implied complexity (if that makes sense). However, depending on which level you investigate genetics (from molecular to population, for example) there are different methodologies that investigate this issue either by e.g. normalizing against environmental aspects (as in many molecular biological/biochemical approaches) or using models to take them into account. The simplistic view being challenged is rarely maintained above bio 101.

 

With regards to your example, it sounds faintly familiar but I only recall a paper from the MPI in cybernetics in which the lead author was actually a physicist, IIRC. While the paper phrased it in terms of noise of sensorimotoric system, the explanation is the same. The question why they looked at it was also from the direction of how we are able to recalibrate our positioning. So that is actually a great example how physics can be used and phrased into a biological context (by using noise analysis in a sensory feedback loop). I actually do not think that they expected a straight line at all (at best as null) but may have expected a more systematic bias. In the end they found that the patterns match up better with noise accumulation.

 

What is true is that biological research sometimes is a bit too empirical and lacking certain theoretical frameworks in many areas. But again, these are the areas were collaborations may shine.

 

 

 

 

 

My anecdotal experience in the Australian system may not be representative, but I saw a number of students either drop out or downgrade to a Masters. My experience in the US differs in that I've seen students pull out early at the prelims stage, or finish (even if they stretch it out the eight years). The difference being that in Australia, you have a hard deadline on your stipend, where as in the US (in my limited experience) the duration for which you can be paid as a grad student is more squishy.

 

That said, there's still a huge range in the quality of work that's considered acceptable to make up a PhD dissertation, even among students with different committees at the same university, and plenty of people out there with a bit of paper and a title who don't necessarily know what they're talking about (and people without them who do).

 

The Dutch and German system is (or used to be) a bit different to both the Australian and the US systems as most of the selection happens before graduation. Only a fraction of the students enter actually graduate with a degree (before Bologna it was a Diplom). Only with that degree a few can actually obtain a PhD. Also, there are generally no grad schools as those in the US. Instead, you have full-time research and the ability to obtain a PhD is solely in the hand of your PhD advisor. There are so-called grad schools but they were not comparable in terms of workload.

Especially younger Profs are letting their students get a PhD eventually, as it may reflect badly on them, if they do not have sufficient graduates from their labs. Often the Diploma time (which is about 6-9 months of research) is used to gauge their ability to eventually obtain a PhD.

Edited March 21, 2014 by CharonY

[jimmydasaint]

In response to the OP, I suspect that the authors know a reasonable amount about the biology of what they are saying. For example:

One experimental way to resolve the nucleus/cytoplasm issue is cross species nuclear transfer to enucleated eggs. This has not proved possible with mammals, but has been successful with fish. Enucleated goldfish eggs transplanted with nuclei from carp eggs develop with the outward appearance of the donor carp, but with a vertebral number (26 to 31) consistent with goldfish (26 to 28) rather than the genomic DNA donor carp (33 to 36). We assume that when two dynamic attractors are placed in a common environment, as in the case of the zygote, that they will “synchronize” as, for example, with Huygens’ clocks. Therefore, we argue that biology can explain inheritance on the basis of a sound foundation in the appropriate physics, without resorting to mechanistic narratives involving genes.

Furthermore, work in the 1970s demonstrated that enucleated HPRT-competent (HPRT is an enzyme whose absence causes the awful Lesch-Nyhan syndrome, an inborn error of metabolism-RL) fibroblasts in vitro could correct HPRT deficiency in fibroblasts with an intact nucleus, by transferring molecules via gap junctions, without the need for protein synthesis. In addition, erythrocytes (red blood cells) dispose of their nuclei at the last stage of differentiation, but retain, for example, the circadian rhythm function for their lifetime.

 

http://blogs.plos.org/dnascience/2014/03/20/challenge-supremacy-dna-genetic-material/

[chadn737]

In response to the OP, I suspect that the authors know a reasonable amount about the biology of what they are saying. For example:

http://blogs.plos.org/dnascience/2014/03/20/challenge-supremacy-dna-genetic-material/

 

They present a handful of cherry-picked examples, but then dismiss a vast body of research spanning 100 years that says the contrary. Even their own examples do not fully support their claim. Consider the goldfish/carp experiment. These fish develop the outward appearance of a carp, the only feature more similar to a goldfish being the vertebrate number. If genetics were irrelevant, then why expect the fish to look like carp at all, in line with their genetics? I am not saying genes are 100% deterministic, no geneticist would, but these physicists have made the opposite mistake of claiming genetics to be all but irrelevant.

 

The HPRT example comes from in vitro cells, not in vivo. What this means is that if you supplement a cell with the correct molecule, that you will get normal function. Of course, but that does not make genetics irrelevant. Thats like saying that genetics is irrelevant in Diabetes because you can supplement the person with Insulin shots. That person still has diabetes.

 

Its called cherry picking your examples and they do so very badly.

[vampares]

The fractal of mmm-human nature. The paper is benign and will not score anyone an executive GLF4. Didn't evolve me. So...I'll evolve myself.

 

The Snowden thing won't fly either, so... Human Nature. Fine stuff.

[hypervalent_iodine]

!

Moderator Note

vampares, stop bringing up the Snowden nonsense in other threads. Stick to the topic without trying to being in your own.