Wikipedia:Reference desk/Archives/Science/2012 May 17

Science desk
< May 16	<< Apr | May | Jun >>	May 18 >

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

May 17

What is the formal definition of a chromosome, or is there one?

I cannot seem to find a definition anywhere, Wikipedia contains conflicting information across different pages, as do Google results. Hoping someone here with a genetics background can enlighten me. Chromosome says that it is DNA complexed with protein, but I've read elsewhere that it is DNA complexed specifically with chromatin. The latter definition would exclude prokaryote circular DNA from being classed as chromosomes then.

Also, in regards to karyotype, I've heard at various times both a chromatid and a pair of homologous chromosomes being referred to as a chromosome in the singular. Humans have 23 pairs of chromosomes, right? So 46 different (heterozygous?) chromosomes in a normal interphase cell? Sorry if I have overcomplicated this, I just can't get my head round it at the moment. Thanks for any clarification in advance -Zynwyx (talk) 09:30, 17 May 2012 (UTC)

The first part is is a very good question, which urges me to do some remedial reading on prokaryote biology - thanks for asking it. As explained in chromosome, the term is used "loosely" to refer to the bacterial chromosome, which is technically a genophore. But when the literature is full of people speaking of the bacterial chromosome, that attempted redefinition seems like a failure.

So what is the criterion to distinguish a natural chromosome from a mere plasmid?

First, let's consider what it's not. A plasmid has an origin of replication, or it couldn't replicate; and the vast majority of the time it contains a partitioning locus (see segrosome), or otherwise strains would be spontaneously cured of it with fairly high frequency. (See [1] for some pretty data) The plasmid apparently is associated with the nucleoid at least the majority of the time (see previous and [2], though I'm a bit tentative about this). The distinction I find in one book is "Probably the best distinguishing characteristic of a plasmid is that it has a more typically plasmid origin of replication with an adjacent gene for a Rep protein rather than a typical chromosome origin with an oriC, along with a dnaA gene and other genes typical of the chromosomal origin of replication".[3]

Problem is, bacterial artificial chromosomes and phage artificial chromosomes don't meet this definition! As explained here, they're based on the F episome (F plasmid), but as our article explains, they do have the Rep gene. Also, lest we have any confusion, these are closed circular constructs which often undergo homologous recombination after injection to a mammalian host to form linear transgene arrays [4] - they are not eukaryotic chromosomes with telomeres and such like yeast artificial chromosomes. So they're called "artificial chromosomes" in the bacterial sense.

So what distinguishes a plasmid from an artificial chromosome? I'm going to guess that much of the distinction is practical: how much stuff you can cram into it. A high copy number plasmid is going to recombine with itself, sometimes lose pieces of the gene, and the ones with the least stuff in them will replicate best. So when you want to shove a lot of stuff into bacteria, you want one of these more well behaved low copy number vectors with more natural-ish performance. And so, in the absence of a very clear definition, we have this sort of continuum from natural chromosome to artificial chromosome to plasmid to the odd random piece of lost circular DNA, without as clear a defining line as we might like. But I might still be just not thinking about something important... Wnt (talk) 15:24, 18 May 2012 (UTC)

-----Thanks for the detailed response! I think genetics has become a victim of the speed of its own success in this respect, as there aren't concrete definitions for loosely thrown about terms. So at least there are genetic differences between a plasmid and a bacterial genophore (in terms of origin of replication, which is good to hear. A chromosome in eukaryotes though, as I have now learnt, can refer to:

• a single strand of DNA, what might commonly be referred to as a chromosome when you look at a karyotype. E.g one member of the homologous pair of chromosomes we inherit from each parent.

• in mitosis, two strands of identical DNA joined at the centromere (the X-shaped things, to put it in very non-technical terms). Each strand is in itself an individual 'chromosome' as by the first definition, and indeed one of the strands is a replica of the other, but in this situation each strand is called a chromatid and the two identical 'chromatids' joined together are known as sister chromatids. When the sister chromatids separate, each one then becomes known as a (daughter) chromosome again.

So this was confusing me when reading up on mitosis/meiosis as I found it hard to keep track of what chromosome/chromatid was going where and so on. I don't personally think it is a good use of a word, much like bringing two identical twins together and calling them one person, but I guess that is personal opinion and its something that is ingrained and unlikely to change. By the wider definition though of simply a large 'section' of genomic DNA (chromatin or not), both situations above are fine (albeit still slightly muddling to me). Another semantic muddle-up I was having was homozygous and homologous. Homozygous refers to when the two 'sister chromatids' are identical in gene sequence and allelic content (barring random mutation). Homologous refers to when two separate chromosomes or two 'chromatids' are the same in terms of what genes they encode for and the sequence of the gene loci, but differing in alleles. So in chromosomal crossover, two 'chromosomes' (which are composed each of two homologous and homozygous 'sister chromatids') come together to form what is known either as a tetrad or bivalent (more confusion...) and then homologous but heterozygous 'chromatids' (ie one chromatid from each X-shaped 'chromosome') recombine to form in total four homologous (but heterozygous to one another) chromatids, which then separate to become four homologous chromosomes. Phew! I'll definitely be thinking more carefully in future in using these terms!! -Zynwyx (talk) 13:25, 19 May 2012 (UTC)

Or maybe not! According to this page the two sister chromatids together can also be called a dyad. Is it correct to call the two sister chromatids (the "X") a chromosome then? -Zynwyx (talk) 13:42, 19 May 2012 (UTC)

I'm not following your reasoning with the last step. If you have 46 pairs of chromosomes, they first replicate their DNA to form 92 chromatids. They're still 46 chromosomes because each has 1 centromere. (That isn't as odd as it sounds because, remember, there's a period when the DNA isn't fully replicated, and you only definitively know that's over after they are starting to be pulled apart) Now the chromosomes pair up to form 23 bivalents, with exchanges between sister chromatids. I'm unaware of people calling them anything but 46 chromosomes and 92 chromatids at this point. After the exchanges, which involve crossing over and thus are incomplete, the bivalents get pulled apart, so you have 23 chromosomes and 46 chromatids in each. Then the chromatids are pulled apart and you have 23 chromosomes and 23 chromatids in each gamete (1n). So 1 chromosome = 1 to 2 chromatids, 1 bivalent = 2 chromosomes, 1 tetrad = 4 chromatids, 1 dyad = 2 chromatids. Unless I fouled up ;) Wnt (talk) 14:38, 19 May 2012 (UTC)

Thanks, that makes a lot more sense. I have been thinking this whole time of a chromosome as being a helix of DNA which holds one allele of a gene, and thus couldn't get round in my mind how the X things are called chromosomes as well. Obviously each X derives from one chromosome that is being replicated, and the replicas are joined at the centre, so it makes a bit more sense in my mind now. So 23 pairs of homologous chromosomes > 46 individual chromosomes > 92 chromatids during mitosis (but still 46 chromosomes). That clears it up in my mind a bit, thanks! -Zynwyx (talk) 08:30, 20 May 2012 (UTC)

Metal loses its magnetic properties when in a liquid state?

I was just reading this, http://en.wikipedia.org/wiki/Electromagnetic_projectile_devices_%28fiction%29#Literature

And it stated that "Later it was shown that molten metal cannot be accelerated by a magnetic field as metal loses its magnetic properties in a molten state, and Clarke admitted his error gracefully." Is that true? In a railgun however, a piece of metal need only be conductive, not magnetic. Would molten lead or an eutectic of lead and bismuth still remain electrically conductive? ScienceApe (talk) 15:28, 17 May 2012 (UTC)

Mercury (element) is a metal in it's liquid state at room temperature, and, I believe, retains electrical conductivity. Perhaps what they mean is that, if significantly perturbed, the liquid will then break into droplets, making it more difficult for an electric charge to pass between them. StuRat (talk) 16:25, 17 May 2012 (UTC)

Mercury is absolutely conductive in its liquid state, see mercury switch for a common electrical use of mercury. --Jayron32 17:23, 17 May 2012 (UTC)

It must be possible because you can buy one. This link is to the Permanent Magnetic Pump (PMP), but the data sheet also mentions an Electromagnetic Pump (EMP). --Heron (talk) 18:35, 17 May 2012 (UTC)

A railgun uses an electric current to generate a magnetic field in the projectile, accelerating it through simple magnetic repulsion, and can use any projectile material that remains conductive during acceleration. A coilgun is essentially a linear motor that accelerates a magnetized projectile, and so requires that the projectile be able to remain magnetic during acceleration. You can (theoretically) fire a liquid metal from a railgun, but not from a coilgun. --Carnildo (talk) 23:04, 17 May 2012 (UTC)

If a nail is heated red hot, as with a propane torch, it reaches its Curie temperature and is no longer attracted by a magnet. Other ferromagnetic metals have varying Curie temperatures. Iron would still be electrically conductive when heated red hot, though its resistance would change.Molten iron and steel would still be conductive, since an electric arc furnace relies of current flow from an electrode to molten metal in the crucible. Metals typically have an increase of resistivity of 1.5 to 2.5 in the liquid form compared to the solid form at the melting point per [5]. Edison (talk) 23:54, 17 May 2012 (UTC)

I would agree with Heron, induced eddy currents would make it possible. An eddy current separator accelerates nonmagnetic metals in the same way. Ssscienccce (talk) 08:14, 18 May 2012 (UTC)

Could the molten metal imagined by Clarke have been above its Curie temperature but contain a suspension of a powdered metal with a higher Curie temperature? 84.209.89.214 (talk) 14:40, 18 May 2012 (UTC)

Yes! A ferrofluid! You can even make one yourself fairly easily. That said, it's useful to understand why something is no longer magnetic in liquid phase. It's essentially the definition of a liquid - the material no longer has any defined ordering of its atoms (or constituent molecules), and ferromagnetism in particular depends on electrons being able to align in some fashion and remain that way. It's a useless suggestion if the atoms are turning and tumbling over each other as happens constantly in a liquid. That said, one could have something putty or gel-like which flows like a liquid but has defined crystal structure at a small scale. This is kinda a cheat, but it might work for some sci fi - a very-much-zoomed-in example would be like a giant tank of those mini-magnetic spheres. They will flow. SamuelRiv (talk) 02:11, 21 May 2012 (UTC)

Space radiation

Lets say you are in outerspace and you are protected from everything (pressure, you have oxygen, temperature, etc) except radiation. About how long would it take before you die from radiation? ScienceApe (talk) 15:46, 17 May 2012 (UTC)

- which part of outer space? Some parts are full of radiation, and you'd die immediately. Some parts have very little, and you'd die of old age. 91.125.207.125 (talk) —Preceding undated comment added 16:03, 17 May 2012 (UTC).

I don't think there's enough radiation, near the Earth, to kill from radiation sickness. There is enough, however, to cause genetic damage and cancer. But these things aren't always fatal, so there would be a decreased life expectancy, not certain death. If you were closer to the Sun, the radiation damage might be more severe.

Note that I'm assuming that you are excluding UV light. If that is included, I'd expect bare skin in space (if somehow protected from the cold and vacuum) to quickly burn, crack, and bleed. Death might occur within hours or days, from dehydration and infection. StuRat (talk) 16:04, 17 May 2012 (UTC)

The article Health threat from cosmic rays may be of interest to you. LukeSurl ^{t c} 18:46, 17 May 2012 (UTC)

Define "radiation". What portions of the electromagnetic spectrum are you including, and what portions are you excluding? --Carnildo (talk) 23:05, 17 May 2012 (UTC)

Quoting from a NASA report (Shielding Strategies for Human Space Exploration dec 1997):

• In prior manned space missions, the GCR have been considered negligible since the mission times were relatively short and the main radiation concern was the very intense SEP events which can rise unexpectedly to high levels, delivering a potentially lethal dose in a few to several hours which could cause death or serious radiation illness over the following few days to few weeks if precautions are not taken [4]. The most intense such event known occurred on August 4–5, 1972 between the Apollo 16 and Apollo 17 missions [5].

GCR = galactic cosmic rays, SEP = Solar Energetic Particles Ssscienccce (talk) 15:21, 18 May 2012 (UTC)

Mario Rabinowitz

I was reading through the Ball lightning article, and it has some stuff about black holes by someone named Mario Rabinowitz. At a first look the Mario Rabinowitz article makes him out to be someone very impressive, but I can only see links to ArXive papers, and I can't see any affiliation with physicists I've actually heard of, or a position at a university or laboratory. Looking at the article's history, it looks like almost all of it was added to wikipedia by people who haven't done other things. So I'm a bit suspicious. I can't find anything worthwhile about this person by searching Google (I find facebook and patents and whitepages and stuff, and copies of the papers). So I'm concerned that a] this person doesn't really exist at all (that the article is a hoax) or b] that this person does exist, but isn't a physicist anyone has heard about (and so maybe should't be on wikipedia). Or is he really a famous physics guy who I've just failed to hear anything about? 91.125.207.125 (talk) —Preceding undated comment added 15:59, 17 May 2012 (UTC).

He's a real person,[6] but it sure looks to me like he's a non-notable person who wrote an article about himself as an autobiography, and the article should be deleted. But WP:AFD would be the place to bring that up, not here. Red Act (talk) 19:29, 17 May 2012 (UTC)

Cooperative lightning? Intelligent grass?

I'm sure this is a very dumb question, but assuming that what this website says is true: When lightning begins to travel downward from a cloud, many objects that have built up a charge emit streamers. This could come from anything such as a blade of grass or a power pole. The first streamer to make contact with the bolt defines the final path the lightning will take. - then how do the ground objects "know" when it's time to start emitting streamers? A simple answer please, for this admittedly ignorant non-scientist. Textorus (talk) 16:03, 17 May 2012 (UTC)

See our article on lightning. Fundamentally, though, positive streamers from the ground emerge for the same reason that negative streamers from the cloud (i.e. the formative lightning strike) emerge -- there's a large electric charge differential present. — Lomn 16:11, 17 May 2012 (UTC)

You might also want to know how gravity works. That is, how does the Earth "know" there is a star 93 million miles away which it should orbit ? Some rather non-intuitive explanations emerge, such as space being curved, or the even weirder gauge boson theory. StuRat (talk) 16:20, 17 May 2012 (UTC)

Now that you mention it, that is a fascinating question, but I'll save that one for another time. (The aether gets knotted up into a rope, maybe, like a yo-yo string?) Textorus (talk) 18:01, 17 May 2012 (UTC)

Lightning is a very complicated electrodynamic phenomenon. In addition to the visible incandescent stream of hot gas that you see, there are also wide-band electromagnetic waves (radio waves), preceding the lightning strike, occurring in tandem with the lightning strike, and coupling with the complicated streams of moving ions and electrons. The radio signals from a lightning strike are often called "sferics." Like all other radio waves, they travel approximately at the speed of light. The actual event of a lightning "striking" may be preceded by a very quick burst of radio-energy; and then as the streamer forms, all sorts of electromagnetic effects start happening and interacting with each other chaotically. The gas gets hot and incandesces, releasing visible light (incandescent light); but the gas is also ionizing and forming an electrically conductive stream, providing a current path, releasing more radio-wave emissions; and of course, the radio-waves emitted will affect air surrounding the lightning streamer. Here's a fairly advanced science web-site: Lightning Modeling, that reviews some of the physics necessary to accurately describe what's happening during a lightning strike. Nimur (talk) 17:10, 17 May 2012 (UTC)

The website is above my pay grade, but it makes sense that there must be some connecting force linking earth and sky. So when conditions are right for a storm, you're saying there's already a lot of ions and electrons moving between the two, invisible to our eyes? (I'm sure I must have learned that in Physics 101 but that was a l-o-n-g time ago.) Textorus (talk) 18:01, 17 May 2012 (UTC)

I think this is actually a very interesting question. The electricity clearly finds a quite specific path. My suspicion is that even in the absence of thunderstorms we are surrounded by an amazing display of static electricity which is simply, most often, too weak for us to see. Certainly I know that during a thunderstorm I can feel frequent little shocks from a mattress if it contains metal springs - sort of the sensation of being first bitten by a mosquito, but of course without the mosquito or subsequent irritation. Sometimes I've ever observed sparks from a window screen though lightning was not nearby. Has anyone ever sought to visualize the wider web of static electricity, or is it simply impossible, or indeed, am I deluded? Wnt (talk) 18:04, 17 May 2012 (UTC)

HAIL Project sought to visualize the wider web of electrodynamics in the atmosphere on massive geographic scales. By monitoring perturbations in the continuous background of electromagnetic signals (specifically, several LORAN transmitters), data was collected to drive a complete realtime model of the ionization and the electromagnetic environment for the continental US. Nimur (talk) 20:58, 17 May 2012 (UTC)

A former colleague of mine attempted to measuring electric fields near clouds by photographing polarization changes through an optical telescope. This is called the Kerr effect, and refers to the change in optical properties of certain materials (like atmospheric air) when exposed to very strong electrostatic fields. I recall thinking the idea was crazy (the signal should be well buried in the noise); but that's why it's research... Nimur (talk) 21:02, 17 May 2012 (UTC)

I can't search videos online at work, but I encourage you to find the ultra slow motion videos of lightning, it's frikken awesome! The "streamers" they talk about are a LOT slower then the speed of light and there are videos of them propegating though the sky, once they "contact" eachother, the lightning bolt actually fires like a flash. It's one of the most incredible natural phenomena on earth I think. Vespine (talk) 22:56, 17 May 2012 (UTC)

And of course, the fine folks in Gainesville, shoot off rockets trailing metal wires. Nimur (talk) 00:33, 18 May 2012 (UTC)

The streamers start due to electrostatic forces: attraction between positive and negative charges. To use an analogy: Imagine a cardboard box with a bunch of nails, screws, iron filings and stuff. If you hold a magnet above it, the objects start moving, pointing in the direction of the magnet. Get the magnet closer and eventually one of the objects will jump up and attach to the magnet, usually pulling others along. Electric charges act in a similar way. Ssscienccce (talk) 08:24, 18 May 2012 (UTC)

Thank you for that plain-English answer. I thought it must be something like that. Textorus (talk) 18:50, 18 May 2012 (UTC)

Of course, there's still the question of exactly why forces can act at a distance, whether electromagnetism, gravity, etc. StuRat (talk) 18:54, 18 May 2012 (UTC)

Where can drugs get in your body? What barriers are there?

I'm aware of the blood-brain barrier but presumably there are plenty of other barriers. Where can orally or intra-venously taken drugs get in your body? Presumably anywhere blood can get, but where is that? Can drugs get inside cells? What about bones? What about your eye lenses? What effects whether they DO get there?

I think that's enough questions for now... although I've got a lot more.

What would be a good place to start looking for the basics of this stuff?

Egg Centric 19:50, 17 May 2012 (UTC)

Hair analysis shows that detectable levels of many drugs can be found in your hair. Long-haired drug users effectively carry a timeline of their drug consumption imprinted in every hair of their head. --Jayron32 20:31, 17 May 2012 (UTC)

Plus, if a man has long hair, can't we just assume he's a drug addict ? :-) StuRat (talk) 18:56, 18 May 2012 (UTC)

An oral drug should get into a cell at least once, in the intestinal epithelium, in order to enter the body. These and many other drugs usually have their effects inside a cell. Sort of an exception are antibiotics, which act inside a cell, but not one of yours! ;) There are exceptions, though - any drug which blocks a cellular receptor, for example. So far as I know any injected drug (well, "biologic") with a name ending in "-mab" (monoclonal antibody) will not get into a cell for meaningful purposes (it might get endocytosed with a receptor and have a trip to the lysosome, but that hardly counts). Vaccines don't get into cells, at least not the old fashioned kind available on the market. There's nothing quite like the blood-brain barrier and even that allows some things to pass. Certainly bones are visited by Fosamax and its ilk. The lens of the eye is a curious case, as it receives sustenance from the aqueous humor from the ciliary body; thus drugs must go by this indirect route; nonetheless they can arrive. For example, acetaminophen overdose can form cataracts in experimental animals receiving the drug systemically;[7] however, this occurs after it is first processed by the liver to form a more toxic metabolite. [8] Wnt (talk) 21:28, 17 May 2012 (UTC)

The place to look for basic information is any introductory pharmacology textbook. Generally speaking in order to get into cells a chemical needs to be lipophilic, meaning capable of dissolving in fats or oils. That's basically the same thing required for a drug to cross the blood-brain barrier. The exception is substances that are transported by active uptake mechanisms. Looie496 (talk) 23:47, 17 May 2012 (UTC)

Thanks all Egg Centric 17:57, 18 May 2012 (UTC)

Note that the skin acts as a barrier to some, but not all, drugs. Hence transdermal patches. StuRat (talk) 18:58, 18 May 2012 (UTC)