CharonY

I see what you mean, and this is also what I meant earlier (though admittedly badly worded, also note that this is not really my field so I may be remembering things inaccurately). Depending on how you test for HWE you can infer different mechanisms for violation thereof. E.g. if you only tested the haplotype frequencies, you would have found that they are not in HWE (I faintly recall some odd methods using Fisher test to ascertain that, although I assume that there must be some kind of regression analyses, or some even more simple models which I simply am not familiar with). Also statistically, not refuting HWE does not automatically mean that the given allele is in HWE (as you cannot accept the null, although sometimes it is done so in literature). And of course (but that is implicitly stated in your example) a given allele may be found in HWE but it may not hold on the population level (i.e. considering the whole genome pool).

I remember that the Gates foundation have put the Cornell lecture online. Maybe a quick google for that?

Normally, the HWE calls for static allele frequencies (often used as a null). Drift and non-random mating even in absence of mutation or selective pressures would violate that. There are (IIRC) certain modifications of testing for HWE that try only to elucidate the effect of fitness influences. In these cases drift and other stochastic events are somehow taken out of the equation. But in the classic sense I am with Ringer, drift violates HWE. Originally drift was possibly not mentioned, however it was implicit in assuming very large populations (which minimize stochastic influences).

There are several research directions going on. Unfortunately biological agents, such as phages are hard to dose and control and often do not give satisfactory results under non-optimum conditions (e.g. clinical trials). Other approaches including e.g. silencing are also at least near-trial, though I am not familiar enough to have an opinion on that. My overall feeling is that at one point or another that resistances are almost unavoidable. The biggest chances are likely vaccines, especially those targeted at essential virulence factors. That, however, requires much more base knowledge on the details of the actual processes. I do not think that any magic bullet will suddenly appear that will solve all the problems.

The laminar flow hood (or biosafety bench) is primarily designed to protect the user from biological contamination. The laminar flow reduces spread of aerosols and works as an air curtain. While the clean air and reduced turbulence also assists in sterile techniques, a zero flow box or even a horizontal would actually be better. In zero flow there is even less air movement (obviously), whereas in horizontal the flow blows contaminants away from the sample (but towards you, hence, not suitable for biosafety purposes). That being said, standard procedure is finding a low flow area (i.e. no passage of people, no door or window or airducts), and work near a burner. The latter is a bit tricky as the burner actually creates turbulence. So sometimes it may make things worse. Depends quite a bit on the type of organisms you got, however.

What is the stoichometric ratio of chlorophyll to protein?

If we look at proximate mechanisms then there is little left for a free will per se (esp. in the libertarian sense). The closer we go in time to the actual execution of a decision the more locked the pathways appear to be. Decisions for movements and other decisions are apparently all prepared in the brain before we become conscious of the actual decision, for instance. However, given the complexity of neuronal activities the predictive power (in a deterministic sense) diminishes as the biological noise overwhelms the data we get. For instance, if we measure brain activities in the ms to second regime it is possible to observe decisions being made even before they become conscious (Libet and later Hynes have done some nice work on it). But move back the timeline further it will become almost impossible to foretell when and which decision is going to be made. The mechanisms however, are still more or less deterministic, just the activities and resulting outcome are complex enough that they appear not to be.

My bad, carry on. missed a post. It is true that on the physical level we do not know all of the details (or even of all the players). There are similaritities to quorum sensing in as much as signal gradients trigger specific cellular responses. Obviously the stratification within biofilm is much looser than in actual tissues. Direct cell-cell interactions play roles in both systems, however. Obviously, in most cases direct neighboring communication is insufficient as it does not integrate enough data to create a larger tissue or form tissue boundaries (though boundary conditions can be conveniently assumed in models).

While we are at it, do not put the hot bottle near your eyes to check whether everything has been dissolved. Superheating is no fun, tight cap or not.

If you mean bacteria that can utilize arsenic instead of phosphorous (GFAJ-1) then I have to tell you that they most likely don't. Finally LC-MS was done and it seems pretty much the end of the story (well, provided it gets published, so far the data is on arxiv and I was not able to get hands on samples myself). If you talk about arsenic resistant and/or reducing bacteria, I would look at the ATCC. There are quite some collected there.

Not everyone is doing that. At least I do not know anyone who does it. What I could imagine is that it is an attempt slow down the heating, but then you may get the same result by using a lower setting (though some may not allow that).

From how the question is phrased it is of no relevance. Essentially you test whether A and B are from the same population. You have to be careful with the null, however. If your null is that B performs the same or is worse, it is a different test than A and B perform the same (which I think is the simpler and hence more likely intended question). For the second question you are essentially a correlation analysis and you have to create the null differently. Think about the alternative hypothesis a bit more.

Garlic also functions as a anticoagulant, which helps preventing thromboses. Of course that has downsides, too, especially if one is prone to ulcers and internal bleeding. On the other hand one has probably eat a lot to have strong effects.

Depends a little bit on how established the methods in your lab are. If results are unexpected I would unleash the whole range of controls to identify the problem.

I tend not to think in terms of complexity as (I think) it is quite a loaded term. As mentioned in other discussion, biology has many imprecise, but useful definitions (such as species, for instance). I just do not feel that "complexity" is a useful term in this context. All organism that we observe today were subjected to evolutionary mechanisms for the same amount of time, as such assigning different level of complexity as a whole is not terribly helpful. If one focuses on only one element then one can clearly distinguish between levels of complexity, but it is not really a reflection of evolutionary reality. For instance, dinosaurs, as a whole, are not more or less complex than existing birds or reptiles. However, if only looking at the ability to fly, birds have improved quite a bit. Also the biosphere is not striving towards increasing complexity. Evolution allows more and more niches to be filled, but as a side process of biological activity certain other niches may be destroyed. One has to go back way, way, way back in time to find a situation where biodiversity was low and many niches unused. But at a certain level you will find changes rather than definite trends. Also, it is dangerous to argue from a post hoc position and draw a line towards where we are today, as with different random event outcome may change quite a bit (i.e. trends built from that perspective are not necessarily true, even if you could observe them).

Here is the thing, once a certain level of complexity has reached, it allows for certain other features that are not successful otherwise (e.g. by gaining access to new niches). However, it does not mean that it will continue. If there was a continuous trend, all "simpler" lifeforms would eventually be replaced, which is certainly not the case. Furthermore, basing complexity on one certain feature (e.g. anatomy) is, in a way, cherry-picking. If I focus e.g. on metabolic abilities instead, a handful of bacteria would be far more complex than e.g. all animals taken together. Also the scaling is skewed, anatomical complexity is not directly represented on the genetical level. This is evidenced by the relatively moderate increase of genes from bacteria (~4k genes) to humans (~20k). Of course one could argue that regulation could be more complex, but then it depends on what element one focusses on. Developmental regulation is of course far more complex in multicellular organism. But then metabolic adaptations and stress responses are far more limited.

Whatever it is, it is not microscopic. I presume taken with a Nikon (DSCN prefix).

What is this assertion based on? Most chemical reactions that more complex organisms are capable of, can also found in larger variety in the prokaryotic realm for instance. And even if that was true it is easy to extrapolate or construct trends from complex data. However, it doe not mean that a trend really exists.

Just out of curiosity and as a sidenote, does anyone know what that means, precisely? It sounds a bit like a truism: eat food and your metabolism goes up, because food is being metabolized. It does not stay that way, of course.

A very good point and biology is full of convenient, but artificial distinctions. However, one could argue that the underlying mechanisms are likely to be somewhat different and possibly more varied. Though the basic idea (i.e. frequency changes over time in dependence of structure/functionality) is similar.

This is a very good point. And if we wanted to critique description of the tour guide in terms of precision one would also have to know the correct wording. Obviously other metrics (i.e. certain statistics) paint a better picture in terms of scientific education. However, without looking at statistics I can think of the movement to teach "alternative theories" to evolution in science classes as detrimental to science education.

Also, education is learning to become educatable.

Have you tried counting live-titers (i.e. plate assays?). Also have you established a good calibration curve between absorbance and actual cell count?

The efficiency is incredibly low, especially if only bacteria are involved. In most cellulose feeding animals a more complex system including flagellates are needed. And then the hole system needs adaptations to the digestive tract after all, to create an environment where cellulose degradation can occur. And even then it is often not terribly efficient. Horses are known to have terrible colics and they are not very efficient feeders. Cows are much more effective, but they have a very complex digestive system to allow that. Chances are that if you introduced (even in relatively high titers) the GMOs, they will be outcompeted very quickly.

I disagree with that. The blackboxing is based on empirical data and very rough statistical inference, i.e. we know that they grow at roughly that particular rate and then expect them to be ready in roughly the same time frame. The key here is that normal biological noise is being suppressed (i.e. using chemostats or standardized batch-techniques, for instance) so that fairly rough models that use little if any biological information and mechanism can give out (simple) parameters with a certain degree of accuracy. The equivalent in economics would look at the same resource distributions and roughly predict flow under the very same conditions and predict the outcome, which is somewhat feasible. But as with biological behavior under realistic conditions the data is usually much, much noisier. Now I challenge you to predict the growth rate of a cell population with a new sugar source with any degree of accuracy. Now with several parameters changed. Now add a stress and try to predict metabolic changes. Now with several varying parameters and a given mutation. Now let us add gradients and microenvironments and complex populations. And how about predicting behavior of single cells? We are not able to predict e.g. the proteome or metabolome changes in an organism upon environmental changes, nor are we able to use that data to accurately predict the current state of a given cell. Predictive biological models are really very very unreliable.