Wikipedia:Reference desk/Archives/Science/2021 December 16

Welcome to the Wikipedia Science Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

December 16

COVID-19 mutations

Most seem to be an amino acid changing at point number so-and-so. 20*19 or 21*20 theoretically possible letter pairs. Are the scariness ceilings of the letter pairs about the same? Are some much less likely to make the news (perhaps ones involving aminos that aren't mentioned often on the original spike gene or the subgene for its binding region? These parts only have 72 to 1273 amino acids) Sagittarian Milky Way (talk) 00:19, 16 December 2021 (UTC)

Amino acids do not mutate. Rather, the genetic sequences coding for amino acids mutate. In the case of SARS-CoV-2, that genetic sequence is in the form of RNA (specifically, positive sense single stranded RNA). I have no idea what you mean by "scariness ceiling," can you clarify that? As for whether there are mutations not likely to make the news, yes, there most likely are many of them. Again, because the genetic sequence is RNA and not amino acids, multiple different (but usually similar) codon sequences can code for the same amino acid. This means that there can be point mutations that don't change the translated amino acid. Those certainly would not make the news. Even when mutations do result in a codon change to a different amino acid, this may not have much of any effect on the virus. If the amino acid is in a part of the spike protein that doesn't interact with Angiotensin-converting enzyme 2 and doesn't change the overall conformation of the spike protein, then the amino acid change would have little, if any effect, and wouldn't make the news. If the mutation was somewhere other than the spike protein, and didn't appreciably change the overall performance of the virus, that also wouldn't make the news. Even when changes occur in interaction regions, they may or may not be enough of a change to be newsworthy. Remember that it is the chemical properties of the amino acid side chains that are important to their roles. If the mutation replaced a valine with a leucine amino acid residue, for example, it is possible that the effect would be minimal since both are aliphatic amino acids with similarly branched structures and similar overall size, and therefore will both behave as similar hydrophobic side chains. That doesn't mean such a mutation will definitely have no effect, but it very well might have no overall effect, and therefore not be newsworthy. These types of changes would essentially become a background "noise" in terms of the overall virus population genome. However, replace a valine with a proline, which tends to add "kink" to the overall protein strand, and we have a different story. --OuroborosCobra (talk) 17:25, 16 December 2021 (UTC)

Amino acids do not mutate. Rather, the genetic sequences coding for amino acids mutate. In the case of SARS-CoV-2, that genetic sequence is in the form of RNA (specifically, positive sense single stranded RNA)

I was aware. A base mutation to a synonym codon doesn't matter as much though, naively it shouldn't matter at all.

scariness

concerningness? among variants a few have enough "concerningness" to enter the variant of concern set but I don't think mutations are binned as systematically

ceiling

highest plausible. If there are say 500 valine codons in the nominal original genome then however bad or not that bad the worst of the 500 VS switches would be for humanity would be VS's "badness ceiling". So maybe the chemically different amino acid pairs have higher ceilings and lower floors? And appear more often on lists of concerning substitutions?

I am aware of the basic process but not the minutiae and was just curious about the substitutions distribution, if "which letter pairs are more likely to be bad" or "can all letter pairs be bad" is not a useful or deep concept in virology and is more like recreational mathematics then I was aware of that possibility. Variants of SARS-CoV-2#Notable missense mutations says 17 simple amino swaps have been notable so far and of the acids swapped to (there's way too few notable mutations so far to count pairs yet) G, H, K, R, S and V appear twice, F, N, the proline you mentioned, Q and Y appear once and the other 9 or 10 amino acids haven't been notable yet (I found that SARS-CoV-2 has all 20 common acids from a paper on one of the proteins but can't easily find evidence that the virus has any selenocysteines). Are some amino acid types overrepresented here? Amino acids with the basic (alkaline) tag on the article are overrepresented but it could just be random chance. Sagittarian Milky Way (talk) 20:55, 16 December 2021 (UTC)

Suppose you have a message consisting of an English text, composed of letters and some other signs such as the space character. Random mutations will generally transform a meaningful message into garbled nonsense, such as i will come → i zill uome. Occasionally, something meaningful will emerge: i will come → i kill some. The new message may be more scary than the original. But it does not make sense to discuss the scariness of individual letters in such messages; the process from letter in a word to the whole word to to the meaning of the word to the meaning of a sentence is too complex to assign meaning to individual letters. It is not different for the process from codon to genetic segment to protein to protein function; see the articles Protein biosynthesis, Protein folding and Protein–protein interaction.  --Lambiam 09:01, 17 December 2021 (UTC)

The subset of the ~1270 spike aminos (teal) that touch the ACE2 is tiny

Yes I'm aware of the very basics, that was my first guess (that the category "bad point substitutions" is circa a random digraph generator) but with amino pairs greatly differing in similarity and the low number of ACE-touching aminos and the 2 cleavage sites in each spike protein presumably bearing only 4 aminos I'm not sure. Coronavirus spike protein#Immunogenicity says "The majority of antibodies from infected individuals target the receptor-binding domain". Sagittarian Milky Way (talk) 21:23, 17 December 2021 (UTC)

I found out that only 72 amino acids are in the receptor-binding motif so the average number of copies of each letter that touch the ACE is 3.6 and the small subset of the 72*19 possible substitutions that aren't like i zill uome might not have equal "mutation badness". Sagittarian Milky Way (talk) 22:04, 17 December 2021 (UTC)

And there's no methionine touching the ACE2 receptor at all (at least in the oldest strain). So I've found (as in new to me) that at least 19 letter pairs can only be possible in the other ~9,700 amino acids. Sagittarian Milky Way (talk) 22:24, 17 December 2021 (UTC)

No histidine or tryptophan either. These 57 letter pairs are "low badness ceiling" candidates. Sagittarian Milky Way (talk) 22:35, 17 December 2021 (UTC)

No methionines in the 223-amimo long receptor binding domain and only two in the outer half of the spike protein (673 amino acids). Sagittarian Milky Way (talk) 23:05, 17 December 2021 (UTC)

There are 519 methionineless aminos in a row from early S1 thru early S2, is that random chance? I suppose when you have 20 chances one slightly surprising result is almost expected. Dec 2019 strain has 168 copies of methionine in its non-structural proteins (7,095 amino acids) so about a dozen would be expected in 519. Sagittarian Milky Way (talk) 00:23, 18 December 2021 (UTC)

How does touching corona of sun mean touching the sun?

Most of websites like this says: NASA Parker Solar Probe 'Touches The Sun' For The First-Time Ever.

But it only touched Sun corona not the Sun. Rizosome (talk) 05:32, 16 December 2021 (UTC)

The corona is part of the sun. --←Baseball Bugs ^{What's up, Doc?} carrots→ 05:40, 16 December 2021 (UTC)

A lot of them put quotes around "touch" or "touches the sun" (such as your link) to indicate that it didn't really touch the surface of the sun. Clarityfiend (talk) 07:43, 16 December 2021 (UTC)

Is our atmosphere part of our planet? --←Baseball Bugs ^{What's up, Doc?} carrots→ 13:38, 16 December 2021 (UTC)

Not sure, and how about the geocorona?  Card Zero  (talk) 18:26, 16 December 2021 (UTC)

The Sun Is A Mass Of Incandescent Gas, it doesn't have a surface per se, in the sense that there is a distinct line we can define as a surface. --Jayron32 15:41, 16 December 2021 (UTC)

This sentence solved my question: It doesn't have a surface per se, in the sense that there is a distinct line we can define as a surface. Rizosome (talk) 02:11, 17 December 2021 (UTC)

Resolved

Rizosome, I'm not entirely sure that that is the best answer to your question though (skip to the last sentence of this post

for its most important point if you dont feel like reading all of it). The Sun's surface would, indeed, seem to be more clearly defined than the surfaces of the gas (& "ice") planets. See Surface of the Sun.

the sun has no solid surface, as the stony planets do. It isn't really 'gas' per se, though. Plasma is its own state of matter. But, with all of that said, I think the best analogy of that of the earth and its atmosphere, including its exosphere. Draw parallels between the sun's surface with earth's surface, and the sun's corona with earth's exosphere. (Also, make no mistake: the solar corona is never considered to be the Sun's surface; difficult or not to clearly define, by all definitions the corona is in regions above/outside the surface. 2600:1702:4960:1DE0:39AF:8AB2:A850:B03D (talk) 18:05, 17 December 2021 (UTC)

As I understand it (which may not be terribly well; caveat lector) the visual "surface" of the Sun is essentially an optical illusion. Light rays emerging from layers of the photosphere and traveling far from vertical get refracted as they pass through less dense layers. If I've done this right in my head, they get bent further from the vertical than they started, which means that to an observer, they appear to come from higher layers than they actually do. However in the higher layers, this effect is less pronounced, and the rays are bent less than the ones from lower down. This means that the light from the various layers tends to bunch up together, and then from our point of view, suddenly stop when you look at large enough angle off the center of the Sun.

That gives the visual appearance of a "surface". But in actual fact there is no discontinuity at that point.

Again, this is my vague understanding; if anyone can elaborate further, or refute me altogether, please have at it. --Trovatore (talk) 21:49, 20 December 2021 (UTC)

If you could magically turn off solar atmosphere refraction the photosphere would still be less than 1 pixel wide even in eyeball-exploding 4K Ultra HD. I've never seen a telephoto sunset with the pink/red chromosphere visible either, though it's naked eye for a few seconds tops during total eclipses (grazing and similar size ratio eclipses excepted). I think you can see stars behind it. Photosphere density is also only ~0.0002-0.0004 times air and plummets to nearly lunar levels in the first few thousandths of a radius above the photosphere. I wonder how much that can refract when Earth's atmosphere makes extra-atmospheric objects c. 0.5-0.6 degrees higher at the horizon. Sagittarian Milky Way (talk) 01:46, 21 December 2021 (UTC)

It's a matter of the scale height of the sun in the area around the photosphere and the mean free pathlength of a photon. The scale height is

$H={\frac {RT}{Mg}}$

with R the gas constant, T the temperature (about 5770K), M the molar mass of a mixture of mostly atomic hydrogen and helium and g the acceleration of gravity. Filling in the numbers, you get a scale height on the order of 100km, which is tiny compared to the radius of the sun. Every time you climb this scale height, the pressure drops a factor 1/e=0.37 and density drops with it (not exactly proportional, as the temperature does some strange things).

The mean free pathlength of a photon, i.e. the distance it can travel on average before it gets scattered, depends on the density of the gas and its composition. Where the gas is ionised (hydrogen is hard to ionise; metals are easy), the free electrons scatter the photons very well, making the mean free pathlength much shorter. Where the photons scatter, the temperature of their blackbody spectrum gets adjusted to the temperature of the gas.

As you get higher in the sun, the density drops, increasing the mean free pathlength of the photons. At some point, the mean free pathlength is 10km, tiny compared to the size of the sun and much smaller than the scale height. After travelling this distance, the density has hardly changed, so the photon is still trapped. But then comes the point where the mean free pathlength equals the scale height. Now the photon can travel 100km. But after it has travelled 100km straight up, density has dropped about 63%, increasing the mean free pathlength to 270km. After travelling 270km straight up, the density has dropped even more, increasing the mean free pathlength to 1500km. Within a few hundred kilometres, the mean free pathlength of the photons increases from only a few kilometres to practically infinite kilometres. From below this layer no light can escape the sun, from above this layer all light escapes at once. Therefore, this layer of last scattering is the photosphere, and the temperature of the radiation escaping from here matches the temperature of the gas in this layer. Above the photosphere, the temperatures of radiation and gas are decoupled.

For photons not travelling straight up, the altitude of last scattering is a bit higher, but even at an angle of a few degrees above horizontal, the difference is no more than a few scale heights. For this reason, we can see a few scale heights deeper into the sun near the centre of the disk than near the edge. Deeper in the sun it's hotter, so the centre of the disk is a bit brighter than the edge.

BTW, I consider a plasma just any type of sufficiently electrically conductive fluid. That makes gas and plasma orthogonal concepts. PiusImpavidus (talk) 11:35, 21 December 2021 (UTC)

Well, I did ask for it. Thanks, SMW and Pius. --Trovatore (talk) 17:52, 21 December 2021 (UTC)

The universe does not care about making everything nice and neat and firmly delineated for a species of apes on one particular speck somewhere. Our Sun is a whole lot of nuclei and electrons bound up by gravity. Per the physical laws of the universe, it gets less dense with distance from its center. Where the Sun "begins" and "ends" is a distinction that we humans ourselves create, to some degree arbitrarily. It's fine if you don't agree with the distinction someone else draws, but there isn't an objective way to say who's "wrong" or "right", unless perhaps someone is being really far-out ("the Sun extends to the center of the Milky Way galaxy"). --47.155.96.47 (talk) 02:04, 23 December 2021 (UTC)

Highest refractive index lens image a human eye can experience

Reference desk/Archives/Science/2021 December 16

GaP (impure)
GaP wafer (electronic device quality)
Properties
Band gap	2.24 eV (indirect, 300K)
Refractive index (nD)	2.964 (10000nm), 3.209 (775nm), 3.590 (500nm), 5.05 (354nm)
Hazards
NFPA 704 (fire diamond)	3 1 0
Except where otherwise noted, data are given for materials in their standard state (at 25 °C [77 °F], 100 kPa). Y verify (what is YN ?) Infobox references

What substance and illuminant would you need? Does anyone make human-usable lenses from it? Are there several answers depending on things like will you accept an SDTV level of image quality or very long wavelengths which would need like a room lit by bright laser diffuser(s) and viewed through a peephole lens of the material? I've read that humans can see up to 1999 nanometers without damage if it was bright enough (the threshold for decent images is likely shorter). The highest I've found in 400-700nm is gallium phosphide, which gets niche use as orangish-red lenses that cost more than sapphire (if this is what electronics grade looks like then lens grade must be pure lol). Or you could send laser(s) through the lens and compare before and after angles if you don't want to diffuse the laser into better lighting, this would work with prisms too. Sagittarian Milky Way (talk) 16:30, 16 December 2021 (UTC)

References

Rutile has a high refractive index 2.609 and can be transparent to light. Most of those high index materials also have a high dispersion and so are not that useful for optical lenses. Gallium phospide, being a semiconductor is pretty opaque at optical wavelengths. I doubt that you can see anything at 1999 nm. Our infrared article says 1,050 nm as a limit. I can see 850nm as used in some optic fibres, but I can't see 1310 nm. Graeme Bartlett (talk) 06:04, 17 December 2021 (UTC)

You can see through lots of semiconductors if the bandgap is high enough. Like diamonds. There is a Czech paper on gallium phosphide lenses for red and infrared light. Maybe you have to be dark-adapted? Does eye aging affect longwave transmission too? A fiber optic would be harder to see than half the visual field blinking or something like that. I could see 940nm remote controls but stopped doing that a long time ago as some people say it's bad for the eyes. Above a certain wavelength multiple infrared photons have to pop in the same vision molecule nearly simultaneously to see anything, and often enough to be distinguishable from the non-zero appearance of pitch dark. Sagittarian Milky Way (talk) 07:24, 17 December 2021 (UTC)

Diamond would count as an insulator at standard conditions. But it does have quite a high refractive index. The high band gap pushes the absorption edge into the ultraviolet. For a semiconductor, the band gap would have to be of low enough energy, so that the average heat energy can promote some electrons to higher energy. But that energy is well down into the infrared, so don't expect much from semiconductors in transparency. Graeme Bartlett (talk) 10:50, 18 December 2021 (UTC)

List of semiconductor materials has it, how many electron volts does something need to be an insulator in standard conditions? Are transparent LED emitters insulators too? The Czech paper says that the 50% transmission level for optical purity gallium phosphide is 0.6 microns which isn't very precise but seems visible enough. The photo of it is red. Sagittarian Milky Way (talk) 12:32, 18 December 2021 (UTC)

Looks to me that the Wikipedia article you quote is just plain wrong. Diamond is an insulator, and used to be widely used as a heat sink in semiconductor lasers because of its very low electrical conductivity and high thermal conductivity. See the BBC here for one citation of this.--Phil Holmes (talk) 17:52, 18 December 2021 (UTC)

See also this Wikipedia link--Phil Holmes (talk) 18:06, 18 December 2021 (UTC)

Should the article explain the temperature and conductivity ranges? And show which conductivity ranges major electronics categories like light emitters and solar panels are in? A phase-diagram looking thing would be succinct. Sagittarian Milky Way (talk) 18:35, 18 December 2021 (UTC)

Do flowers wait to be pollinated?

Do some flowers wait to be pollinated before they fall off? Does this explain why the flowers on some indoor plants, like orchids, last so long? Some of mine last for years. Thanks. 86.187.161.127 (talk) 20:30, 16 December 2021 (UTC)

The article section at Orchidaceae#Pollination explains quite well that pollination opportunities in orchids can be scarce, so the flowers indeed remain receptive for long periods. Outdoor species may be similar but wind and rain can cause pollination even when insects are not present, so keeping plants indoors will tend to encourage lasting flowering. Mike Turnbull (talk) 22:31, 16 December 2021 (UTC)