Politics Discussion Thread (Heated Arguing Warning)

[greenline]

There isn't a single politician on the left who wants to "subsidize moving Mexico's population" or literally replace our constitution with Sweden's or whatever; those are gross perversions of legitimate arguments for immigration reform and universal healthcare.

Perversions? What is the purpose of repeatedly granting amnesty, of repeteadly not coming up with any solutions and instead trying to lower barriers to immigration, of trying to openly defy immigration laws if not to make it easier for them to come over? What was Bernie Sander's campaign if not an attempt to move the European model of socialism over here?

Have whatever opinion you want on these policies, but don't deny they exist

[MJW (ya that one)]

Ps. This is Apolyton c.2001. I await the implosion with glee, abandon and dunkin' donuts. And beer, but maybe that does need sayin'.

Asher did the dirty work...

[Huinesoron]

I'm talking about new ways of making the old drugs.  We should define a drug by a certain chemical formula, crystal structure, excipient, and purity.  Maybe define some standard laboratory tests as well, which it needs to pass, to mimic blood-absorption rates or whatever else I'm overlooking, having gone into a field where I'm allowed to use my knowledge to reduce costs.  Instead we define it by 'went through these manufacturing steps'.  Basically we assume that chemistry is not a field of knowledge, that medicine is magic and only by waving the magic wand in the exact same manner as before can we treat people.

I'm mostly staying out of this, but as a professional chemist this is directly relevant to me, so I felt the need to step in. Yes, chemistry is a field of knowledge. And one thing it tells us is that analytical chemistry is... hard. You have to know exactly what you're looking for in order to find it.

[Long, pointless, detailed ramble removed for your mental health. ]

In lieu of that, an example: my company makes a fairly simple product - just one small molecule containing less than 20 atoms, in a solution that contains a couple of other components. We make it at three locations, using the exact same starting materials and method. We then test two things - strength, and free acid. And... over the past several years, we have not once managed to get the ratio between those numbers to be the same across all three sites.

Why? We don't know. It works in the lab! But when you step up to the full-scale reactors, it just... doesn't. The engineers blame the chemists, the chemists blame the maintenance team, they blame the computers, IT blame everyone... it's a nightmare.

Luckily, this is a well-studied substance, and it still falls within the scope of the regulations and what's been previously studied. If it was an expensive drug, that wouldn't be the case. How much variance are you willing to allow? Does your medicine need to be within 10% of 'the same' before you'll take it? 5%? 20%? Because what it will never be is identical.

So yeah, you define anything that can't be simply identified - and complex organics like pharmaceuticals absolutely come under this - by its manufacturing method. And if you want to change that, you get to do all the identification work over to demonstrate what Europe at least calls 'technical equivalence' - ie, you have to prove it's the same stuff, not just say 'yup, looks about right'.

Because otherwise you accidentally introduce a thousand times the normal levels of vanadium. Or you have 5% of the wrong stereoisomer in your product. Or your powder has a different structure, dissolves faster, and gives your patients an overdose. Or your molecule folds slightly differently, and therefore affects the wrong protein. Or, or, or...

Chemistry is a field of knowledge. Medicine is not magic. But doing them right is hard, and accidentally killing someone is so, so easy...

[darrelljs]

Darrell please stop looking like a decent human being that I would agree with. It's disconcerting.

Didn't I once compare you to Dick Cheney ?

Ps. This is Apolyton c.2001. I await the implosion with glee, abandon and dunkin' donuts. And beer, but maybe that does need sayin'.

I was around for that!  RB remains the one corner of the Internet where people can have civil, reasoned debates on very charged topics without it degenerating into ad hominem rants.  Well, mostly .  I put it down to the fact we are generally older, and smarter .

Darrell

[Krill]

And I think Dick Cheney was described in a positive manner...

[Bacchus]

Experiment in action: if RB can stand up to factionalism, there is at least hope for the wider world.

I'd like to hear the ramble on analytical chemistry though -- in the common understanding of science there is a real dominance of physics as an exemplary science, and understanding something can be equated with having a tractable analytical model built up from first principles and fundamental particles that predicts observable behavior. Of course, that's not at all how science in general works -- chemistry and biology in particular, but we don't hear enough about that. It's a pet grievance of mine, and on occasion I rise up to defend phenomenological methods (important for economics, which is my discipline), but I really value hearing from scientists in these fields.

[Huinesoron]

I'd like to hear the ramble on analytical chemistry though...

On your head be it.

I'll start with a real-world example: one day, I was handed seven bags of white powder. (I shall pause here for the customary jokes... done? Good.) I was told that they were taken from seven different containers out on the plant, but that for whatever reason, they were entirely unlabelled.

I knew that there were two different white powders commonly found on site - one being zeolite (it's a sodium-aluminium-silicon mineral), the other aluminium hydrate. The easiest way to distinguish between the two is that only one of them dissolves in acid - but of course, I didn't know for sure that they couldn't be anything else.

Instead, I turned to a neat bit of kit called an X-Ray Fluorescence spectrometer. Essentially, it shines x-rays on the sample, and detects what elements are present by the light they emit in response. It won't detect anything lighter than sodium, and different elements show up at different strengths, but it's a good first approximation. I stuck the first bag under, and got back readings of 55% aluminium, 25% silicon, and 3% sodium - so that's your zeolite, then. Fine. And all of the first six bags matched that.

The trouble came with the seventh bag. It looked just like the others, but gave a completely different reading: 2.4% aluminium, 2.4% silicon, 2% sodium, and a whopping 92% chlorine. Chlorine? That's not zeolite, but it's certainly not aluminium hydrate either!

At this point I turned to the What-Is-This machine, which... ha! I wish. Instead, I had to make a guess at what it might be, and devise a test to check. I tried dissolving it in water, and found that it was very soluble, which meant it was (probably) some kind of salt - but what sort of salt? 'Salt' just means ____ Chloride, and we've had four or five different salts on site.

One possibility was ammonium chloride (which contains nitrogen and hydrogen, neither of which the instrument could detect), but that stuff stinks to high heaven, so no. Magnesium or potassium would have shown up on the initial results, and the sodium level was far too low... but on the other hand, the instrument is notoriously bad at detecting sodium.

It would have been nice and scientific to find another way to detect elemental sodium, but we don't have one in the lab. What we do have is samples of sodium chloride (AKA table salt). Pop one under the x-ray source, and sure enough: 2% sodium, 92% chlorine. It's salt.

That's how much trouble I had identifying a simple metal salt, containing two atoms per molecule. Now step that up to some complex organic drug, with dozens of carbon atoms that need to be in exactly the right place... I'm not a pharmaceutical chemist, but here's the tests I'd want to perform to be sure a new production method was making the same material:

• X-ray diffraction. If it's crystalline (and stable as a solid), we can use this to get an idea of the structure of the molecule. We'll get a diffraction pattern, which is like working out where a handful of pebbles fell in a pond by looking at the ripples. If we're lucky enough to make a large crystal, we can use a different instrument to get a map of the electron density in the molecule.
  
• Testing for every functional group that should be in the substance - and every one that shouldn't be, but might have gotten in there. Do we have a double bond? An alcohol? An ester? We need to check for all of these.
  
• Level of purity. It's great that we've detected the double-bond we were after, but how sure are we that there's not 1% of the single-bond variant? That would be a completely different molecule. I'd probably try testing whether boiling or freezing separated the components, and work from there.
  
• Particle size. This will affect how fast your drug dissolves once taken, so yes, it's important. Luckily, there's a nice bit of kit to test this by shining lasers through it, wonderful stuff.
  
• Trace metal testing, down to parts per billion. How would you feel about receiving ten times the expected dose of arsenic? No, me neither. So we'll want to test for... well, every metal on the periodic table. I know my company has a £10,000+ instrument to do this, that still couldn't get down to the resolution I'd want for a medicine.
  
• Stereoisomers. Some molecules can 'lock' into two different forms, which can be very biologically different - it's commonly said that thalidomide is only dangerous in one of its two stereoisomers. I... don't actually know how to test for these, to be honest. They're chemically identical, so you need to delve into biology for that.
  

And after all that? I still wouldn't be comfortable saying it was definitely the same stuff. 90%, maybe higher if I could get that more precise x-ray result at the beginning. How do you feel about a nine-out-of-ten chance that your medicine will do what it says on the tin?

Basically, chemistry is going to need an entirely separate test for everything you want to observe as either present or absent. Physics lets you stick a cloud chamber on your particle accelerator and spend a year analysing the tracks you detect. Chemistry gives you a rack of glassware and a cupboard full of chemicals and tells you to figure it out. And the only way to be absolutely sure what something will do to a biological system is to stick it into one.

[scooter]

That is truly a top-tier threadjack. This forum rules sometimes.

[darrelljs]

What scooter said .

Could an electron microscope be used to do an atom by atom structural analysis?  I guess if there is variance from molecule to molecule you'd have to take a statistically significant sample size...there's also the problem of this being Physics coming to the rescue of Chemistry ...

Darrell

[darrelljs]

And I think Dick Cheney was described in a positive manner...

Just to be clear, your theory is I made an analog to you that was positive ?

Darrell