Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Maths becomes biology's magic number

bbc.co.uk

Post reply on HN
Print view View original

31–38 of 38 posts

Re: Maths becomes biology's magic number

#31

Post by vacri »

llamaz wrote: post #19
vacri wrote: post #17

Earlier quoted context omitted.

I am very aware about statistical methods originating in mathematics - I did statistics every semester as part of my biology degree. Indeed, half the reason why the mathematics department existed at my university was to support the biological sciences. You may be interested to know that my point is that you don't need mathematical proofs to have a pragmatic, beneficial effect - and while the OP is sitting there feeli…

I think he/she is just being playful and humorous. But since I'm studying math along side EE, I feel like I should comment about the role (or lack thereof) of proof in practical applications. MRIs fall into the field of electrical engineering, and EEs generally don't understand the math they use. e.g. when working with medical imaging the radon transform is used, which would require a post graduate mathematical educa…

Part of my issue here is that saying I mentioned. I've experienced first-hand a couple of times where a mathematician tells me something is or isn't possible due to theory X or whatever, and then when practically applied, they've been wrong. So when someone is smug due to something in their head that doesn't have a practical demonstration, I'm suspicious from the outset.

Maths itself may be pure and correct, but it isn't the maths being smug at the dinner party :)

Re: Maths becomes biology's magic number

#32

Post by inopinatus »

vacri wrote: post #31
llamaz wrote: post #19

Earlier quoted context omitted.

I think he/she is just being playful and humorous. But since I'm studying math along side EE, I feel like I should comment about the role (or lack thereof) of proof in practical applications. MRIs fall into the field of electrical engineering, and EEs generally don't understand the math they use. e.g. when working with medical imaging the radon transform is used, which would require a post graduate mathematical educa…

Part of my issue here is that saying I mentioned. I've experienced first-hand a couple of times where a mathematician tells me something is or isn't possible due to theory X or whatever, and then when practically applied, they've been wrong. So when someone is smug due to something in their head that doesn't have a practical demonstration, I'm suspicious from the outset. Maths itself may be pure and correct, but it i…

The trick is to first raise the topic of "alternative" therapies so that you can all have a go at being smug. It then becomes an amusing dinner-party contest in who can be the smuggest and no-one is offended, or at least, no-one present.

Re: Maths becomes biology's magic number

#34

Post by hilop »

GarrisonPrime wrote: post #4

As a physician who was a mathematics major in undergrad, I completely agree. An astonishing amount of biology is rather mechanistic, systematic, and logical. I feel my math (and minor programming) experience trained my brain in such a way as to understand complex biological interactions more intuitively. I'm not bragging here, but while I was actually understanding the theories behind what we were learning, on a fund…

Common wisdom is that for most STEM and professional graduate degrees, the best undergraduate preparation is a math degree and a minor in the related field. If you do a major in any science or social science, have the time is wasted on tiptoeing around the math the prof is afraid you can't handle.

Exceptions: lab/field work aren't covered by math class (do these as your minor and senior year) , and theoretical physics/chem do the math -- but you are better off doing a year of math and a year of accelerated whatever, than doing intro whatever and then advanced whatever.

Re: Maths becomes biology's magic number

#35

Post by GarrisonPrime »

xaa wrote: post #30
GarrisonPrime wrote: post #4

As a physician who was a mathematics major in undergrad, I completely agree. An astonishing amount of biology is rather mechanistic, systematic, and logical. I feel my math (and minor programming) experience trained my brain in such a way as to understand complex biological interactions more intuitively. I'm not bragging here, but while I was actually understanding the theories behind what we were learning, on a fund…

Bioinformatician here. All my formal training is in biology, but I've spent most of my research hours learning CS, statistics, etc, rather than biology. > An astonishing amount of biology is rather mechanistic, systematic, and logical. I disagree vehemently. Or, to be more precise, it is very likely that this is true in reality, but in terms of the dominant way it's viewed in both clinical and research practice, it's…

I didn't meant to imply biological systems were simple A-->B boxes or machines with perfectly predictable actions and reactions. I worded my comment poorly.

You are quite right that it's a dynamic, multi-faceted reality. I think the blank looks are saying that it's unreasonable to ask because it's impossible to know; all they're doing is trying to tease out at least one connection. Hopefully, none of them seriously think they've hit the nail on the head and have found the one true mechanistic answer.

Re: Maths becomes biology's magic number

#36

Post by GarrisonPrime »

hilop wrote: post #34
GarrisonPrime wrote: post #4

As a physician who was a mathematics major in undergrad, I completely agree. An astonishing amount of biology is rather mechanistic, systematic, and logical. I feel my math (and minor programming) experience trained my brain in such a way as to understand complex biological interactions more intuitively. I'm not bragging here, but while I was actually understanding the theories behind what we were learning, on a fund…

Common wisdom is that for most STEM and professional graduate degrees, the best undergraduate preparation is a math degree and a minor in the related field. If you do a major in any science or social science, have the time is wasted on tiptoeing around the math the prof is afraid you can't handle. Exceptions: lab/field work aren't covered by math class (do these as your minor and senior year) , and theoretical physic…

I agree with this assessment.

I've given dozens of presentations to my fellow hospital docs, and by far the most attended and appreciated are those dealing with proper use and interpretation of statistics.

Re: Maths becomes biology's magic number

#37

Post by GarrisonPrime »

qud wrote: post #15
GarrisonPrime wrote: post #4

As a physician who was a mathematics major in undergrad, I completely agree. An astonishing amount of biology is rather mechanistic, systematic, and logical. I feel my math (and minor programming) experience trained my brain in such a way as to understand complex biological interactions more intuitively. I'm not bragging here, but while I was actually understanding the theories behind what we were learning, on a fund…

How did you transition from a Maths/Programming undergrad to doing medicine? I'm doing Computer engineering and plan to do go into medicine later, any tips?

Medicine was always my goal, so I decided I'd use my undergrad time as an opportunity to try something else I was always interested in and which could provide me with some radically different perspective compared to the biology/chemistry folks. (And would also serve as a backup if I failed to get into medical school...)

My transition was easy. The required prerequisite college courses for admission to med school (two semesters biology, two physical chemistry, two organic chemistry, two physics, one basic calculus) were more than sufficient to get me through my med school courses. I took a few extras too (anatomy/physiology, molecular biology, and genetics) but I feel those were overkill and might not have been necessary. At least they were fun.

I think the biggest part though was that I sought out some experiences in the medical field. This demonstrated to the admissions committees that my career goals weren't a lark and I knew what I was getting into. I did a 6-week summer mini-med-school camp, shadowed physicians in the ER a couple of times, and volunteered in a medical records department for a semester. When I didn't get into med school the first time, I spent the interim year getting my Certified Nursing Assistant license and working in an assisted living facility and a hospital's locked psychiatric ward. Fun times.

Good luck.

Re: Maths becomes biology's magic number

#38

Post by xaa »

GarrisonPrime wrote: post #35
xaa wrote: post #30

Earlier quoted context omitted.

Bioinformatician here. All my formal training is in biology, but I've spent most of my research hours learning CS, statistics, etc, rather than biology. > An astonishing amount of biology is rather mechanistic, systematic, and logical. I disagree vehemently. Or, to be more precise, it is very likely that this is true in reality, but in terms of the dominant way it's viewed in both clinical and research practice, it's…

I didn't meant to imply biological systems were simple A-->B boxes or machines with perfectly predictable actions and reactions. I worded my comment poorly. You are quite right that it's a dynamic, multi-faceted reality. I think the blank looks are saying that it's unreasonable to ask because it's impossible to know; all they're doing is trying to tease out at least one connection. Hopefully, none of them seriously t…

> one true mechanistic answer

Well yes, if they thought this, they would likely be wrong. But that isn't what I was suggesting. The answer I would be looking for would be something like "A accounts for approximately 20%/50%/90% of the variance of B in context C". Clearly many or most biological entities have more than one input and output.

My problem isn't that there's complexity, it's that we don't/can't/don't even try to quantify and robustly model it. The reason why I feel it is such a problem is that without quantification of effect sizes, (and obviously these effect sizes would be probabilistic and perhaps context-dependent), we cannot really predict much of anything.

Another way of making this point is to consider the sort of pathway or signal transduction diagram you often see in Nature Reviews, etc. Then imagine for a moment that you are looking at a electric circuit diagram instead of a biological diagram. You have a resistor here connected to that capacitor there, etc...

What's missing? Resistances. Capacitances. More generally, numbers or functions specifying the behavior of the nodes and edges conditional on their inputs. Such an electric circuit diagram would be almost entirely useless to a person who wanted to predict the behavior of the circuit under different conditions.

I guess I just object to the statement (and you are far from the only one to have implied it, if indeed that's part of what you were saying) that we have a mechanistic understanding of X, if we cannot predict the behavior of X conditional on perturbations, even approximately.

For me, this was and is a major stumbling block in my ongoing journey to learn biology. I am often shown such diagrams and told "this is how it works", as if these diagrams had any explanatory or predictive power. I am perpetually confused about how and why people think they do.

N.B. I admit they have SOME explanatory/predictive power, but it seems to me to be very close to zero. The only remotely quantitative aspects of these diagrams are the stimulatory/inhibitory "arrowheads", indicating a positive or negative correlation, and missing edges, indicating presumed conditional independence. These diagrams also lack any temporal information.

Post reply on HN