Hacker News Reader

Comment by toasterlovin | Hacker News Reader

I think the best big picture explanation of why this is important is this: our ability to feed and care for humanity is dependent on biology. Genomes are the source code for biology. Reading the source code is the first step to understanding the source code. Understanding the source code will allow us to build better expressions of biology (those being both ourselves and the animals and plants we depend on).

View on HN

I hope you won't take it personally because you're far from the only one to say it (and I would have refrained from posting if I hadn't seen people taking this further in this thread), but I really wish people stopped using the "DNA is the source code of your body" analogy. It is used way more than what's it's actually worth, even though I can see how that analogy can be seductive. The structure of a program's source code is nothing like the way DNA encodes and propagates information. Even if you add a bunch of adjectives such as 'self-modifying', 'full of GOTOs', 'nondeterministic', 'live monkeypatched', and thus twist the picture many people have of the way source code translates into execution, it would still be misleading. We are a product of both our genome and environment, at the same time, all the time. Not one then the other. Or, if you will, the source code is its own executable. It doesn't make sense because the analogy doesn't either.

And that's not even taking into account non-DNA things that encode and propagate information, such as microbiota or anything linked to epigenetics.

And even more so, a product of our cells and RNA soup and so on. There's a truly huge operating system that DNA operates in. I think of DNA, when I think of it as all in a coding way, as a top-level script on top of enormous machinery. Like a switch on a giant factory, that says 'man' when you flip it one way and 'mouse' when flipped the other. Does the switch encode what it means to be a man? a mouse? No.

Is it better to say DNA is input data and epigenetics is the actual application?

Or is it opposite?

DNA is not input data, it is simply the source code. Epigenetics is the state/memory. The environment is the input data.

It’s actually a great analogy. Analogies aren’t supposed to be exact representations; they are supposed to allow you to quickly communicate a decent approximation of what something is by relying on preexisting understanding of how another thing works. The source code analogy does that well.

And, by the way, running programs are also the result of their source code and the environment in which they execute. It’s just that the environments which computer programs run in are very homogenous and predictable. But flip some bits in memory or have a user thrash the UI and you’ll see that a running program interacts with its environment in much the same way as a biological organism.

I never said it was a bad analogy, I even used a rather positive word ('seductive'). It is overrated, though, and it does stop making sense sooner than many people think. What I'm really saying is that relying on this analogy to understand how DNA works in relation to organism development, function, etc. won't get you past a very cursory understanding of it. It may be enough for many people but:

-I like to think that HN is filled with people who take interest in many things, especially science-related subjects, and would hope that many of them on this site would like to know more than an analogy-based understanding of genomics

-For better or for worse, the recent bloom and advances in genetics and biology in general have taken over the modern world and been hailed (probably with reason) by all kinds of circles, from the media to governments or the tech world, as a promising new era that's going to revolutionize our understanding of life, disease, what it means to be human, cognition, what have you. It does feel pretty cool as an insider to know that you're working in the 'hot new thing' but a side effect of that is, I have seen people develop a pretty bizarre fascination with DNA, whether it be their own or other people's. And it's not something that can be pinpointed to entirely rational reasons. I don't want to go off tangents any further, but let's say it can't hurt to occasionnally remind people what DNA is and what it isn't, what it's like and what it's definitely not like. I dislike pedants as much as the next person but seeing so many (presumably very educated) people take that analogy in the same thread made me think that the reminder wasn't that out of place.

I disagree, DNA is pretty much the source code. any differential DNA expression are due to histone modification, proteins, or RNA (and rarely CpG DNA modification in humans) so DNA in the purest form is indeed the source code.

Source code for what, though? How about pre-natal influences in the womb, does this mean the womb is a sort of compiler? What about the structure of DNA itself, known to influence gene expression levels? What about the microbiota, which is arguably more 'you' than your own body in terms of raw cell count? What about the phages found in the microbiota (the phageome)? Why are they not part of the so-called source code?

Source code/machine code, I am just interchanging but in this analogy they are equivalent. Pre-natal influences change the epigenetics, i.e. "on" the genome. The structure of DNA is almost completely determined by histones, which are proteins that DNA wrap around, so once again, not the DNA. Microbiota affect the environment of the cells, meaning it is an input to the cell, not the source code. Any phages affecting the microbiota will affect the input to the cell. These factors aren't considered source code simply because it only affects the execution of the cell's source code.

The source code for the organism.

Yeah, but the question is: having the source code in the form of a bunch of files (each containing some routines) OR having it in the form of a bag of routines, is that a significant difference?

(To see the analogy, substitute file = chromosome, and routine = gene)

Great explanation, thanks.

Reply on news.ycombinator.com