Maths Becomes Biology's Magic Number (bbc.com)

In the middle of a discussion about the pros and cons of statins, Sir Rory Collins, the head of clinical trials at Oxford University, noted that If you want a career in medicine these days you're better off studying mathematics or computing than biology. A report on BBC adds: It is a nice one-liner, but I didn't think much more about it until a few days later, when I found myself sitting in a press conference to mark the launch of a new initiative on cancer. Rubbing shoulders on the panel with the director of the Institute of Cancer Research, Professor Paul Workman, was a scientist I didn't recognise, but it soon became clear this was exactly what Sir Rory had had in mind. Dr Andrea Sottoriva is an astrophysicist. He has spent much of his career searching for Neutrinos -- the elusive sub-atomic particles created by the fusion of elements in stars like our sun -- at the bottom of the ocean, and analysing the results of atom smashing experiments with the Large Hadron Collider at Cern in Geneva. "My background is in computer science, particularly as it applies to particle physics," he told me when we met at the ICR's laboratories in Sutton. So why cancer? The answer can be summed up in two words: big data. What Dr Sottoriva brings to the fight against cancer is the expertise in mathematical modelling needed to mine the vast treasure trove of data the information revolution has brought to medicine. "The exciting thing is that we can apply all the new analytical techniques we've developed in physics to biology," he says. "So we have all these new quantitative technologies that allow us to process an enormous amount of data, and all of a sudden we can start to apply that to implement the paradigm of physics in biology."

Comments

F!ve F!ngers

[Anonymous Coward]

Tequ!!!a.

Anyone?

Re: F!ve F!ngers

[Anonymous Coward]

Let's talk about Hillary Clinton

Re: Math vs Maths

[Anonymous Coward]

Ah, but Maths is 1\5 the betterer.

Re:Math vs Maths

[Dogtanian]

In the US, Mathematics is abbreviated as "Math" (this article is tagged with "math"), not "Maths".

I always felt sorry that you guys only had the one math.

I thought of starting a fundraiser to buy you another mathematic. :-P

COMPUTER SCIENCE?? O RLY?

[Anonymous Coward]

"My background is in computer science, particularly as it applies to particle physics," he told me...

Hey slashdot, that doesn't mean he fixed computers and changed toner cartridges at the particle physics wing, and it doesn't mean he codes "apps" for particle physics. Finally, slashdot has a summary where actual, legitimate computer science is mentioned and it really is referencing computer science done by a bone fide computer scientist. Wait long enough, you see anything.

Re: COMPUTER SCIENCE?? O RLY?

[Anonymous Coward]

If you think computer science has dick to do with astrophysics has dick to do with particle physics you're as clueless as the doofus this "article" rambles about.

The Graduate (2016 version)

[Anonymous Coward]

Son, let me give you a tip... data science.

..."all of a sudden we can start"...

[turkeydance]

Sir, you are so behind.

Re:..."all of a sudden we can start"...

[Anonymous Coward]

Heh, that was my thought as well. But then I remembered I'm a civilian.

A bit of an exaggeration

[Anonymous Coward]

You'll never end up being the guy designing the experiments or a CSO if you learn math/comp sci instead of biology. You may be hired by those guys to do some work, but that's about it.

Re:Maths, like 36,000

[Anonymous Coward]

Same as what Roger Ailes and Bill Cosby did.

Hospitals and Supercomputers

[Anonymous Coward]

This subject is very hardcore. Just to provide for the needs of surgery, imaging and modelling for better patient outcomes requires a peta-scale supercomputer at the top of the hospital. The models are also complicated as they should not become useless once the patient is standing or walking.

Clobbering With Statistics

[Bing Tsher E]

Clobbering reality with statistics is what you do when you don't have a theory to work with.

It's kind of un-scientific, but it puts all that equipment to 'good use.'

Re:Clobbering With Statistics

[methano]

This guy shouldn't be modded down. This is the first intelligent comment on this article that I've come across so far. This computation cure for cancer crap is just that, crap.

Re:Clobbering With Statistics

[Anonymous Coward]

You hit the nail right on the head!

Re:Clobbering With Statistics

[LostMyBeaver]

I think the mistake we often tend to make is to believe that it is either one side or the other. Computing a cure for cancer is very likely a hopeless approach. On the other hand if we can at some point understand enough about DNA to identify cancerous anomalies and target them through custom tailored retroviruses or nano-tech, I figure... sure why not?

That said, I have been "volunteering" from time to time at a university's biochem department with regards to code optimization. I do this in exchange for lab access so I can learn learn a little about biochemistry. At this point what I've learned is that we really don't know anything at all about biochemistry and instead of focusing all our attention into developing tools that could maybe allow us to actually learn about it, we prefer these insane studies of protein folding an such.

I don't necessarily agree with the original article or how it was written in such a way to sensationalize instead of inform. I think the whole Plato/Socrates conversational thing is entertaining a times, but has very little value outside of philosophy and Hollywood. I do however agree with the sentiment suggesting that there is great value in getting an education that would allow you to make valuable contributions to the study of medicine by taking a less traditional approach.

Of course, I could just be speaking out of the side of my own ass. I like the idea of making improvements to scanning tunneling technology to possibly allow full mapping of a human cell. Then focusing on observing all the molecular interactions that explain the purpose of each part of the cell. Biology labs are almost always completely full of pretty white equipment that looks really really expensive. They even have fancy looking centrifuges.... which is a machine which spins stuff.... around in circles... and it probably cost more than my car (a BMW i3). If I as a computer nerd needed such a thing, I would get a power supply, a mosfet, a motor, an arduino and maybe an IR transmitter/receiver for good luck. Total cost... $100. Biology labs should be located in the same building as machine shops and electronic and mechanical engineers.

Re:Clobbering With Statistics

[Anonymous Coward]

Medical "science" is largely behavioral cause and effect studies. To figure out everything about a single living cell does is too complicated.
That does not stop big pharma from selling poorly researched snake oil at exorbitant prices.
Many diseases have no known causes yet medications for some are sold, with known bad side effects and high cost. Can it really be called science when little or nothing is known? How is medical "science" any better than the bullshit homeopathy?

Re:Clobbering With Statistics

[xtsigs]

Clobbering reality with statistics is what you do when you don't have a theory to work with.

It's kind of un-scientific, but it puts all that equipment to 'good use.'

Right. Statistical analysis has been useless in physics, why should we expect any better when applied to biology?

Of course, some people actually think science is about determining probability in order to predict the future. Those in the know understand science is about making up cool facts that others can believe without evidence, like religion. We don't need even one math, so the idea of maths is just silly.

(Sometimes I wish Slashdot had a desightful rating.)

Re: Clobbering With Statistics

[Anonymous Coward]

You have no idea what you're talking about. Statistical analysis is often the only way to actually understand the large data sets involved in both biological studies and properly set up experiments. Stats are almost universally the backbone of proof in biological papers. Only a tech engineer would think that you can get meaningful data by studying a single piece of biology. That's just not how science works. Burden of proof maybe replication but the initial justification for a theory always involves math and stats. Biology even gets its own number, e.
I mean I was specifically told in college that the best thing to have as a biologist was coding skills. That doesn't mean you can become a biologist by learning code (rather more likely that the biologist learns some code on the side, so that they don't have stupid ideas like op). The point of the article is not that math is a replacement for science, just that the science is (as italeays has been) reliant on math and its subset computer science.
The guy working at the lab who doesn't get protein folding doesn't get it either. This is the general problem with science, people looking at one part of another without understanding how the whole thing works. Winds up being an argument between math and what, no math? Totally inane to any real scientist in any field.

Re:Clobbering With Statistics

[Anonymous Coward]

Graduate student in biochemistry here, curious as to what you mean when you say "instead of focusing all our attention into developing tools that could maybe allow us to actually learn about it, we prefer these insane studies of protein folding an [sic] such". Understanding how proteins fold seems pretty important to me if we're ever to get a proper understanding of enzymatic catalysis, for example. I agree that there's a lot to learn still in biochemistry, but we do know some stuff. Metabolism is pretty decently mapped out, for instance (http://biochemical-pathways.com/).

As for the crazy cost of all the equipment, sure, you can DIY most of the stuff, especially "simpler" things like a centrifuge. But what happens when an undergrad comes into the lab and spins a few bottles without balancing the rotor? Will your centrifuge stop? How many revolutions can your rotor take before it succumbs to mechanical failure from the g-forces? Who's responsible for it if it blows up and takes out an eye? Or the lab? (https://www.aiha.org/get-involved/VolunteerGroups/LabHSCommittee/Incident%20Pages/Lab-Safety-Centrifuge-Explosions-Incidents.aspx) A lot of the cost goes into making sure things work safely. A table-top microcentrifuge with refrigeration is probably around $4k, and a preparative centrifuge the size of a washing machine for spinning down ~6L of culture is about $40k (just under the cost of an i3). Mass production of consumer goods really drives things down.

Re:Clobbering With Statistics

[Anonymous Coward]

Einstein would like to offer you a drink sometimes. All that quantum non-sense and such is getting into nerves of every scientist who gets refused of a nice game of dice with their God.

If it reduces the count of absurd results

[Anonymous Coward]

published, I am all for it. Every day we see articles with studies which cannot be statistically valid being flaunted. It's no wonder two thirds of them cannot be reproduced. If they only applied the analysis required by a first-year physics or chemistry student the world would be a lot nicer place.

Biology? Neutrinos?

[Anonymous Coward]

What should I study? ROFL. How about learning what is what.

Well, duh.

[karlandtanya]

Without math, you don't have science, you're just collecting anecdotes.
Really--I don't know what further to say. I'm just floored this is even a subject for discussion.

Re:Well, duh.

[guises]

This hasn't always been the case. Michael Faraday had no formal degree and used hardly any math and yet contributed a great deal. I just watched a lecture series by him recently (reccomended) and was impressed by how much he was able to demonstrate through nothing but rigorous qualitative experimentation. I kinda have this impression, which I know to be completely false, that everyone prior to the modern era were total idiots who ascribed all natural phenomena to humorous vapors and spirits and the mumbling of witch doctors. To be able to learn something about the physical world from someone who's been dead for 150 years is somewhat revelatory.

Of course, most all of what can be learned that way has been learned that way. So you're not completely off base to say that you can't do science without math if you're talking about contributing to the sum of human knowledge, but a person who learns a thing through rigorous experimentation and application of the scientific method is still doing science, even if what they discover is already known to the broader scientific community.

Re:Well, duh.

[LostMyBeaver]

And it's more important to take the next step of identifying what is and isn't mathematics. Chemistry applies extensive use of mathematics though I've often questioned the underlying principles underlying their means of representing objects and effects. I seem to feel as if there must be massive amount of contradicting methods to describe molecular behavior. This in itself is alright, but have we actually observed this behavior and have we corrected the chemical mathematics to properly represent these observations?

In addition, why is it that we use high school chemistry books today that are based on the same mathematics taught before the invention of the electron microscope. Did we actually understand molecular behavior so well that what we guessed before we had things like scanning tunneling microscopes was so accurate that we don't have to rewrite the books? Or are we just trying to observe only what will prove what we already know?

Re:Well, duh.

[jandersen]

To be able to learn something about the physical world from someone who's been dead for 150 years is somewhat revelatory.

Well, look at some the names that are prominent in science, especially mathematics:

Euclid, ~2000 years ago, the father of, well, Euclidean geometry
Isaac Newton, ~400 years ago, prominent contributor to classical mechanics and differential calculus
Gauss, ~350 years ago, major contributor to just about anything, not least differential geometry
etc

In fact, most of the mathematics and physics you study as an undergraduate at university is at least 100 years old (apart from linear algebra, which is surprisingly young); the only major revolutions since then were general relativity and quantum mechanics, and that is close to a century ago now. What we have been doing since can be described as gap-filling, by comparison. Very important gap-filling, with immense value to all areas of modern life, but the next revolution in science is not going to come until we find a way to unify GR and QM.

Re:Well, duh.

[K. S. Kyosuke]

I kinda have this impression, which I know to be completely false, that everyone prior to the modern era were total idiots who ascribed all natural phenomena to humorous vapors and spirits and the mumbling of witch doctors.

Well, it may be false, but since the "modern era" is considered to have started sometime around 1600, it doesn't include Faraday in any case, so even if it were true, it wouldn't exempt him.

Re:Well, duh.

[K. S. Kyosuke]

Sorry, I meant "the pre-modern era doesn't include Faraday", of course.

Re:Well, duh.

[K. S. Kyosuke]

Modern logic is also surprisingly young. I'd include that in mathematics for sure.

Re:Well, duh.

[xtsigs]

was impressed by how much he was able to demonstrate through nothing but rigorous qualitative experimentation.

My Aunt Erma can telepathically communicate with her cat. Through rigorous qualitative experimentation, I have determined she can call her cat to her just using her mental powers. At least, it seems that way, perhaps more than 50% of the time. Maybe it just happens occasionally. No need to run the numbers, though. Without all those bothersome "statistics," I can tell you that Aunt Erma's cat telepathy is a sure thing because when the cat come and curls up on her lap, Aunt Erma says, "You heard me calling you, didn't you, you wonderful little fur ball."

Good observation and imagination are critical to the scientific process, but they are just the beginning.

Re:Well, duh.

[karlandtanya]

"Maths" is a lot more than arithmetic (A rat in Tommy's house...).
The most important mathematical discipline in science is simple Boolean logic that (at least at one time) was taught as part of the freshman high school math curriculum. The tools of logical thought and formal deduction rather than "hand waving" explanations are the *first* requirements.

More to the point of the article...
Scientists generally organize themselves into a couple of different disciplines--simply because the skill sets (and technical requirements) tend to diverge:
Statistical analysis and mathematical modeling (the subject of the article) are extremely important to both.
The job of the theoretician is to produce a descriptive, predictive, and testable mathematical model and study that model. And to describe experiments and predict what the results might be *if* the model and the hypotheses it's built on are representative of reality.
The job of the experimentalist is to determine whether the model is actually consistent with reality.

Take as a timely example the theoretician's prediction of gravity waves, and around a hundred years later, the experimentalist's observation of same.
Makes a nice ringtone...and a nice conversation starter--when folks ask "what's that noise", you can explain how science takes a lot of persistence and work.

Of course, sometimes experimentalist invalidates the theoretician's model; this is when some of my old professors told me "I was writing Science Fiction...didn't mean to, but the results are in...". Of course you have the theoretician who refuses to submit to experiment, but that doesn't mean they're all charlatans.

Re:Well, duh.

[AthanasiusKircher]

I kinda have this impression, which I know to be completely false, that everyone prior to the modern era were total idiots who ascribed all natural phenomena to humorous vapors and spirits and the mumbling of witch doctors.

This is the fault of both bad science teaching and bad history teaching. The history of science is generally taught as a barrage of "cherry-picked" historical anecdotes that make it look like a series of brilliant unerring people who always seemed to find their way progressing to the next advance. We rarely teach the failures in any depth, certainly not acknowledging the reasons why many learned people used to believe in different models or ideas. When we do bring up some failure -- like phlogiston, to take a prominent example -- it's often just to laugh at it. "Oh how quaint and ridiculous people were back then!"

Meanwhile, general historians often no longer receive in-depth education in science and math, so they're ill-equipped to tackle specialized stuff like the history of science. They may be used to questioning "narratives of progress" or Whiggish history in other intellectual areas, but they don't have the expertise to critique the stories about science.

For years, I've believed the best way to improve this would be to teach the Copernican controversy in some depth. No, not the stereotyped version of the Galileo affair where Science (now endowed with a capital "S") confronts "Ignorant Religion" and triumphs. I mean the real version when in 1633 -- when Galileo was on trial -- nobody had incontrovertible empirical evidence for the earth going around the sun. (That would have to wait until 1727, when James Bradley's chance observation of stellar aberration finally put heliocentrism on empirical footing.) In fact, throughout the 1600s, there were loads of people -- not just Galileo -- looking for things like parallax and Coriolis forces and other evidence of the earth's motion and not finding them. (These were not clearly observed until the 1800s.)

So, contrary to the Galileo stereotype, it was a really tough situation for scientists to figure out what was going on in the heavens in the 1600s. Once Kepler came up with his ellipses (something Galileo rejected, preferring the perfection of circles, a la Ptolemy), there was at least something mathematically more elegant in favor of the heliocentric model. And by the time you have people like Newton coming along, trying to continue to do math in a geocentric (whether Ptolemaic or Tychonic) universe had become increasingly tedious.

If you really want to know how learned people were trying to tackle this problem in the 1600s, it might be good to look at something like Riccioli's 1651 compilation of 126 arguments concerning the motion of the earth (49 in favor of the Copernican hypothesis, 77 against) as part of his 1500-page astronomical treatise. Unfortunately, they've never been translated to my knowledge from the original Latin, but you can read an English summary here.

It's really amazing the sort of reasoned debate that was going on about issues like this. Lots of experimental evidence brought to bear, etc. If we had people who were training to become scientists have a look at stuff like this, they might get a much greater perspective and respect for earlier attempts to produce scientific knowledge... much better than fairytales about Galileo, or Newton getting hit on the head by an apple or whatever.

And, frankly, it might be fodder for discussion of modern speculative theoretical physics, which also has headed toward divorcing itself from empirical verification (at least not in the near future for many theories) in pursuit of mathematical elegance. Newflash to those folks -- some parts of physics for most of its history has depended on major speculative theories with little empirical support. It might be instructive to look back at how earlier generations handled this problem.

Re:Well, duh.

[Anonymous Coward]

Faraday is a curious case. He was very much the experimentalist, but his work led to some of the biggest breakthroughs in mathematics and mathematical physics. Maxwell took his idea of fields and expressed it with mathematical rigour (at the time most physical theories appealed to aethers or obscure mechanical analogies). Mathematicians like Oliver Heaviside took it further and made Faraday's concept of vectors (and vector fields) rigorous (Maxwell had used quaternions instead of vectors). Which in turn led to, well, modern mathematical physics and maths.

Re:Well, duh.

[guises]

Unfortunately, they've never been translated to my knowledge from the original Latin, but you can read an English summary here [arxiv.org].

Hey, that's pretty neat. Thanks. Never even realized arXiv had a "history and philosophy of physics" section.

Re:Well, duh.

[Hognoxious]

why is it that we use high school chemistry books today that are based on the same mathematics taught before the invention of the electron microscope.

What do you think has changed? Which parts are obsolete?

Re:Well, duh.

[Anonymous Coward]

I was the "computer guy" in a medical research facility for a while. In the pressure to get grants, the urge was there to "fudge" (i.e. falsify results usually be throwing out things that didn't fit for one reason or another). The lead researcher would not allow it. He always emphasized that we can learn more from our failures than our successes. I wish that was taught in schools as it applies to just about everything in life.

Maths is not a number

[Anonymous Coward]

It's like, all the numbers, if you think about it.

statistic

[Anonymous Coward]

Reminds me of the quote 'The death of one man is a tragedy. The death of millions is a statistic.'

multi discipline

[Anonymous Coward]

Good to see multi discipline efforts on this fight. How many times have we all heard of people with problems that can be solved with some programming or data mining. If you've got data, talk to a computer programmer!

Purity

[Anonymous Coward]

https://xkcd.com/435/ [xkcd.com]

Re:Purity

[cinnamon colbert]

mod up

Re: And how many maths are there?

[Anonymous Coward]

*please to doing the needful*
FTFY

nonsense

[cinnamon colbert]

I was a graduate student under a very famous biologist
and he often remarked, of all the physicists who have gone into biology, only one or two have made a substantial contribution - their minds don't work the right way
True, we need math for clinical trials, and omics, but for biology you need a feeling for the organism

Re:nonsense

[RightwingNutjob]

Was this EO Wilson? If so, he was an idiot. If he wasn't EO Wilson, he was still an idiot.

Re:nonsense

[Anonymous Coward]

There are probably relatively few physicists doing biology against people who started with biology, so it's likely a smaller number will make a substantial contribution. I'll bet when normalized, the percentages of physicists/non-physicists in biology making substantial contributions are close to being equal.

This is a job for maths!

Re:nonsense

[Anonymous Coward]

As someone who works at the Broad Institute and knows people from physics and finance who have gone into biology, I can say beyond a doubt that the OP's famous biologist must have either not read many papers, not understood how research works now, or was just not thinking much when that comment was made. I don't see how researchers in fields like genetics can get by without at least some programming/analysis/statistics skills - that's what the field is turning into.

Re:nonsense

[Anonymous Coward]

I grew up in a small town in the northern Midwest. And when I went away to college in a big city in the Notheast I was puzzled to hear people describe themselves as "Asian" rather than "Chinese". I mean, what was the difference?

But then, years later after the housing bubble popped, I had to leave the USA to find work - and ended up spending six years living and working in Southeast Asia. Well, long story short, Asia is a really big place - a whole lot more than just China, obviously. And now when someone invokes a generalization about "Chinese" people, I just shake my head and smile. There are a whole lot of different people in that part of the world.

So, anyway, I've also spent long enough doing various biomedical research, and genomics in particular, that when someone makes a generalization about biomedical research I also just shake my head and smile.

There's all kinds of biomedical research. Some of it requires deep mathematical knowledge and some of it requires deep biological knowledge. Some of it even requires both. :) Bottom line, there's huge amounts of work to be done making amazing new discoveries - with room for everyone from a whole variety of backgrounds.

But also a word of caution. I've seen mathematicians and physicists with disdain for biology churn out utter worthless biomedical research because they ignored basic biological knowledge that anyone with an undergraduate degree in biology would know. And I've also seen biologists churn out worthless biomedical research because they couldn't be bothered to learn basic math and physics.

In my experience, if you ever find yourself looking down your nose at the biomedical research community with a haughty sense that they're only stuck because they're ignorant and stupid - well, it's a good time to take a long hard look in the mirror while doing a Google search for Dunning and Kruger. :)

Re:nonsense

[K. S. Kyosuke]

But then, years later after the housing bubble popped, I had to leave the USA to find work - and ended up spending six years living and working in Southeast Asia. Well, long story short, Asia is a really big place - a whole lot more than just China, obviously. And now when someone invokes a generalization about "Chinese" people, I just shake my head and smile. There are a whole lot of different people in that part of the world.

Unsurprisingly, the same thing goes for Europe, Africa, South America...

Re:nonsense

[Anonymous Coward]

My girlfriend's neuroscience lab at the NIH has exactly one researcher who studied neuroscience. Everybody else was either math, physics or computer science.

history repeats

[pcr_teacher]

Physicists have been doing this for a long time. See Max Delbruck and the phage group in the 40's and 50's
https://en.wikipedia.org/wiki/...

Re:history repeats

[bosef1]

Scientist and saint?

Consent isn't something women can actually do

[Anonymous Coward]

The entire concept of "consent" is extremely patriarchal.

Women don't really know what they want... this information is just as unknowable to them as the sensation of pregnancy is unknowable to a man.

Allowing them to vote was and is totally insane.

Re:Consent isn't something women can actually do

[Anonymous Coward]

When women read what you wrote, I'm pretty sure they 100% knew they wanted to slog you in the face and kick you while you're down.

Just like a Trump voter can't distinguish fiction from fact ... it's just as unknowable to them as the rainbow to a dog.

Re:Consent isn't something women can actually do

[Anonymous Coward]

When women read what you wrote, I'm pretty sure they 100% knew they wanted to slog you in the face and kick you while you're down.

95%+ of women are masochists physically and psychologically, you freaking virgin.

Re:Consent isn't something women can actually do

[LostMyBeaver]

Have you actually met a women or are you basing this entirely on what you've seen on pornhub and xvidoes?

Finance quants?

[Anonymous Coward]

Perhaps that's too optimistic of an idea, but if some working in finance can do a useful reconversion, they'd put talent to better use.

Re:Math vs Maths

[superwiz]

The linked article is on BBC. And Britain has already adopted the Indian usage of "maths."

Re:Math vs Maths

[K. S. Kyosuke]

Clearly, mathematika is all Greek to you.

Re:Math vs Maths

[K. S. Kyosuke]

EtymologIcally, it is. But foreign words often dissociate themselves from their original morphology.

Re:Math vs Maths

[K. S. Kyosuke]

No, that's why they need you to go to sleep.

Re:Math vs Maths

[mrbester]

The -a ending implies a plural...

Re:Math vs Maths

[Hognoxious]

Would it make sense to have a mathematic? Would you put it on the same shelf as your economic?

Re:Math vs Maths

[K. S. Kyosuke]

Of course it does. That's why it works.

A Case History In Scientific Method

[srwood]

B. F. Skinner's exposition of the appropriate approach to conducting science: http://courses.umass.edu/psyc241/Skinneracasehist.pdf

Re:Biology? Neutrinos?

[K. S. Kyosuke]

Why not study newt-rhinos while you're at it?

Re:Math vs Maths

[Anonymous Coward]

Yes, right next to the books on datas processings and biologies.