I’ve heard of PCR before (polymerase chain reaction) where more copies of a DNA sample can be rapidly made. If the problem was that the quantity of blood that Theranos uses is too small, why wasn’t PCR used/ (if it was) why didn’t it work?
Also if I’m completely misunderstanding PCR, if someone could for that too, I’d appreciate it, thank you!
In: 148
Counterpoint: if someone knew that it could work, like really could work, they would be swimming in money instead of posting on Reddit.
So the reason why they take so many blood is because each tests take a little amount and they add ups. PCR only copy nucleotide, and right now it is impossible to copy everything in the blood matrix.
I don’t remember the specific claims of the Theranos machine, however *in general* you can use very small amounts of sample (relatively) to measure a lot of things. I work with a lot of immunoassay machines, so that’s what I can talk about.
There are a lot of machines and technologies today that have shrunk the amount of sample needed, and some have shrunk the size of the machine. The Gyrolab uses approximately 1 microliter of blood plasma or serum, or whatever liquid matrix. Roche Cobas units use fairly small amounts, they require somewhere around 100 microliters in the sample cup but take some fraction of that (been a while since I worked with them). Both machines are quite large, though if there was a need to save space you could definitely shrink by some amount.
There are also quite small machines where all the magic happens on a cartridge or plate you load into it, and the machine exists only to read out a signal and/or pump fluids. A lot of work has been done in this area to do multiplexing, which is measuring a lot of different things at the same time using the same sample, but there are limits to how you can do that, and the more you want to measure simultaneously the larger the machine becomes.
So you’re unlikely to ever have a benchtop machine the size of a toaster which can measure the number of things Theranos were saying they could, but you can definitely shrink the current technology down some. But the sample size of a drop of blood is unlikely to ever be possible, as you run up against mathematical limits like the number of molecules of what you want to measure being undetectable at that volume. As it is you can do a lot with a few milliliters of blood, it just has to go to a room full of machines.
I don’t want to say anything is “impossible” but if they were able to deliver what they promised it would have been worth like 4 different Nobel prizes for apparently making half of our processes for DNA analysis and stuff obsolete.
It would be like Steve Jobs introducing the iphone and casually throwing in there that it never needed to be recharged.
Theranos was a scam it didn’t work because it was never meant to, thats the hole point of scams. Fooling people into believing something works/is good when in reality, it doesn’t/never did. Like Ponzi schemes for example, the person doing the scheme knows its a scheme from the beginning but are able to make moves in order to make it all seem legit. If that man sounding blonde who pulled this off were able to continue fooling investors and everyone else, she’d still be doing it
Sampling size matters. If I scooped a cup of water out of the ocean, counted the whales in the cup, then multiplied by cups in the ocean, you would not have an accurate count of whales.
There isn’t enough “stuff” in the sample. To find stuff, you need a lot stuff to look for.
It’s like looking for a single needle in a haystack. The more needles, the more likely you will fine one.
SOURCE: Did some time in bioinformatics.
There were tons of impossibilities with how the product would work, but the simplest issue is:
Imagine looking for a four leaf clover. You can usually find them in a big field of clovers. But if you only had a small patch of grass to look for one, then can you really say that it four leaf clovers don’t exist?
Theranos take such a small sample of blood that it’s near physically impossible ( mathematically improbable) that you’d find whatever antibody/chemical markers of the disease you’re looking for. Some diseases are easily found, but others have such small traces that you’d need a lot more blood to test.
Ugh I was such a big enthusiast about this when I first read about/started hearing about it.
There’s definitely a margin of succession with the technology, but it’s almost like if Apple came out with the iPhone as their first product but really it was just an iPod. They promised more than what their technology was capable of.
IMO they should have started with a nutrient panel test, or really anything that could be tested at a smaller scale. They tried to run before they could walk which was their downfall but the idea was beautiful
Many people have explained why it couldn’t work, so I’ll explain how it could have (possibly) worked.
Imagine we want to build a machine to measure a person’s height but we can’t directly do so. What we can do instead is take data like a person’s shoe size, their heart rate and their skin temperature to infer their height indirectly.
Now, obviously a doctor who needed to know your exact height to make a decision on whether or not to operate would simply measure your height. But if you were just monitoring your height on a regular basis down at the local Walgreen’s, our inference would probably be a decent enough guess to judge whether we needed to go through with the more complicated and invasive direct measurement of height.
Indeed, this is essentially what doctors are doing with actual medical tests. They’re not directly measuring health conditions so much as inferring health conditions indirectly from imperfect measures.
That being said, the device I’m theorizing about above wasn’t what Theranos was trying to build. The device I’m proposing above is powered by solid principles of data analysis. Their device appears to have been powered by magic pixie dust.
I’ve heard of PCR before (polymerase chain reaction) where more copies of a DNA sample can be rapidly made. If the problem was that the quantity of blood that Theranos uses is too small, why wasn’t PCR used/ (if it was) why didn’t it work?
Also if I’m completely misunderstanding PCR, if someone could for that too, I’d appreciate it, thank you!
In: 148
Counterpoint: if someone knew that it could work, like really could work, they would be swimming in money instead of posting on Reddit.
So the reason why they take so many blood is because each tests take a little amount and they add ups. PCR only copy nucleotide, and right now it is impossible to copy everything in the blood matrix.
I don’t remember the specific claims of the Theranos machine, however *in general* you can use very small amounts of sample (relatively) to measure a lot of things. I work with a lot of immunoassay machines, so that’s what I can talk about.
There are a lot of machines and technologies today that have shrunk the amount of sample needed, and some have shrunk the size of the machine. The Gyrolab uses approximately 1 microliter of blood plasma or serum, or whatever liquid matrix. Roche Cobas units use fairly small amounts, they require somewhere around 100 microliters in the sample cup but take some fraction of that (been a while since I worked with them). Both machines are quite large, though if there was a need to save space you could definitely shrink by some amount.
There are also quite small machines where all the magic happens on a cartridge or plate you load into it, and the machine exists only to read out a signal and/or pump fluids. A lot of work has been done in this area to do multiplexing, which is measuring a lot of different things at the same time using the same sample, but there are limits to how you can do that, and the more you want to measure simultaneously the larger the machine becomes.
So you’re unlikely to ever have a benchtop machine the size of a toaster which can measure the number of things Theranos were saying they could, but you can definitely shrink the current technology down some. But the sample size of a drop of blood is unlikely to ever be possible, as you run up against mathematical limits like the number of molecules of what you want to measure being undetectable at that volume. As it is you can do a lot with a few milliliters of blood, it just has to go to a room full of machines.
I don’t want to say anything is “impossible” but if they were able to deliver what they promised it would have been worth like 4 different Nobel prizes for apparently making half of our processes for DNA analysis and stuff obsolete.
It would be like Steve Jobs introducing the iphone and casually throwing in there that it never needed to be recharged.
Theranos was a scam it didn’t work because it was never meant to, thats the hole point of scams. Fooling people into believing something works/is good when in reality, it doesn’t/never did. Like Ponzi schemes for example, the person doing the scheme knows its a scheme from the beginning but are able to make moves in order to make it all seem legit. If that man sounding blonde who pulled this off were able to continue fooling investors and everyone else, she’d still be doing it
Sampling size matters. If I scooped a cup of water out of the ocean, counted the whales in the cup, then multiplied by cups in the ocean, you would not have an accurate count of whales.
There isn’t enough “stuff” in the sample. To find stuff, you need a lot stuff to look for.
It’s like looking for a single needle in a haystack. The more needles, the more likely you will fine one.
SOURCE: Did some time in bioinformatics.
There were tons of impossibilities with how the product would work, but the simplest issue is:
Imagine looking for a four leaf clover. You can usually find them in a big field of clovers. But if you only had a small patch of grass to look for one, then can you really say that it four leaf clovers don’t exist?
Theranos take such a small sample of blood that it’s near physically impossible ( mathematically improbable) that you’d find whatever antibody/chemical markers of the disease you’re looking for. Some diseases are easily found, but others have such small traces that you’d need a lot more blood to test.
Ugh I was such a big enthusiast about this when I first read about/started hearing about it.
There’s definitely a margin of succession with the technology, but it’s almost like if Apple came out with the iPhone as their first product but really it was just an iPod. They promised more than what their technology was capable of.
IMO they should have started with a nutrient panel test, or really anything that could be tested at a smaller scale. They tried to run before they could walk which was their downfall but the idea was beautiful
Many people have explained why it couldn’t work, so I’ll explain how it could have (possibly) worked.
Imagine we want to build a machine to measure a person’s height but we can’t directly do so. What we can do instead is take data like a person’s shoe size, their heart rate and their skin temperature to infer their height indirectly.
Now, obviously a doctor who needed to know your exact height to make a decision on whether or not to operate would simply measure your height. But if you were just monitoring your height on a regular basis down at the local Walgreen’s, our inference would probably be a decent enough guess to judge whether we needed to go through with the more complicated and invasive direct measurement of height.
Indeed, this is essentially what doctors are doing with actual medical tests. They’re not directly measuring health conditions so much as inferring health conditions indirectly from imperfect measures.
That being said, the device I’m theorizing about above wasn’t what Theranos was trying to build. The device I’m proposing above is powered by solid principles of data analysis. Their device appears to have been powered by magic pixie dust.
Latest Answers