4 ms·

Because evolution doesn't work on a species level, it works on a gene level. Genes that help propagate themselves spread more. A cancer causing gene would almos

Because evolution doesn't work on a species level, it works on a gene level. Genes that help propagate themselves spread more. A cancer causing gene would almost certainly be selected against, vs a gene that caused you to live longer and have more children - and spread more copies of itself. The reason people die is not because it's better for evolution. It's because evolution simply doesn't care about maximizing life expectancy. Most organisms die long before they get cancer, so the fitness of a gene that prevents cancer is pretty small.

dekhn 10y ago
I take issue with your first statement- evolution certainly works at the species level (in the sense that population genetics is a thing). It does appear there is a population fitness that is selected for when two desirable phenotypes can't be accomodated in a single individual. Also, "Most organisms die long before they get cancer, so the fitness of a gene that prevents cancer is pretty small." ignores the fact that there are genes preventing cancer (tumor suppressors), and those genes are under active selection.
- rcthompson 10y ago
  The term "tumor suppressor" doesn't necessarily mean that the gene exists for the sole purpose of preventing cancer. It just means that the protein's normal function happens to prevent cells from becoming cancerous. Many DNA-repair proteins fall into this category, for example, simply because repairing errors in a cell's DNA reduces the probability of that cell accumulating cancer-causing mutations. And yet, DNA-repair proteins pre-date the evolution of multicellular organisms, so they cannot possibly have evolved as a response to cancer. In practice, the term "tumor suppressor" generally means a gene for which recessive loss of function mutations are associated with cancer (i.e. you have to lose function of both copies of the gene in the same cell before there's an effect), while "oncogene" means a gene for which dominant gain of function mutations are associated with cancer (i.e. mutation in one copy of the gene is sufficient to have an effect). Neither term implies anything about how the gene evolved or whether selection due to cancer was a factor in the evolution of the gene.
  - dekhn 10y ago
    While most of what you say is generally consistent with mainstream science (including my copy of Cancer by Weinberg, AKA, "the bible"), it neither refutes my point, nor provides any support for the parent poster's claim. Also, "DNA-repair proteins pre-date the evolution of multicellular organisms, so they cannot possibly have evolved as a response to cancer." is not a valid argument. There are eukaryotic DNA repair proteins not found in bacteria, for which evolutionary evidence shows they evolved more recently. A number of such genes aren't just DNA repair, but cell cycle controllers that respond to DNA damage and coordinate complex activity. Whether they are "sole purpose of preventing cancer", is a very deep question, and goes to the heart of the nature of cancer. For example, naked mole rats either never or very rarely get cancer; current thinking is that hyaluronase they produce has a protective effect. Did the genes that make enzymes that produce hyaluronase evolve "specifically to prevent cancer", or did they play other roles and just incidentally provide cancer protection? Those are challenging hypotheses to express and prove either way. I'm sure you're familiar with the Hallmarks of Cancer.
- Houshalter 10y ago
  >evolution certainly works at the species level (in the sense that population genetics is a thing). That's not at the species level. That's, at the very best, at the level of a very small group. And most of those theories have generally been discredited - it only works if the beneficial effect on the group is sufficiently large, see http://lesswrong.com/lw/kw/the_tragedy_of_group_selectionism/ http://lesswrong.com/lw/kw/the_tragedy_of_group_selectionism... >there are genes preventing cancer (tumor suppressors), and those genes are under active selection. That doesn't contradict what I said. Those genes work well enough to prevent cancer in young organisms, but they are clearly not enough to stop all cancer. Why hasn't evolution evolved away all cancer? Because even if there was a gene that could decrease cancer risk by an additional 1%, it wouldn't actually increase fitness that much. A 1% decrease in risk is small on it's own, and then it only affects the 1% of organisms that haven't already died of other things. Preventing cancer is just not in evolution's priorities - at least not past a certain point.
  - dekhn 10y ago
    I concede I was probably inaccurate in my statement about evolution and the species level; note, however, that I disagree with people who believe that the gene is the unit of evolutionary selection (I would refine it to say "functional region" rather than gene because I believe that mutations in enhancers and other transcriptional activators play a bigger role in evolution that mainstream science). On to your second reply: I basically agree. We're mostly arguing about where that certain point is. Problem is: cancer is a disease that is closely tied to many necessary functional parts of multicellular organisms. In a sense cancer is what gets people who weren't got by something more acute earlier, and probably evolution is addressing the more acute issues with a higher priority.
untilHellbanned 10y ago
I think you are arguing on two different levels. I'm a professor biologist and though it strikes me as a pretty indirect mechanism (arguing the effect size could be small), I see some merit to the commenter's idea.
- rcthompson 10y ago
  The current hypothesis is that humans have a longer lifespan because having living grandparents conferred a survival advantage over having only parents. Arguing that elderly relatives confer a disadvantage would seem to directly contradict that hypothesis, unless somehow grandparents were beneficial but great grandparents were not.
  - untilHellbanned 10y ago
    A current hypothesis. There is always room for more ideas.
wyldfire 10y ago
> Because evolution doesn't work on a species level, it works on a gene level. Genes that help propagate themselves spread more. I don't think that's necessarily the case. If my descendants were able to reproduce more because of my gene that reaped me before I was a drain on their resources, that gene could be inherited by them. IIRC I vaguely recall something about male homosexuality being correlated with female siblings' increased offspring count. It's a stretch to say that they'd be caused by the same gene but it would be an interesting phenomenon which might be a net benefit to the species.

dekhn 10y ago

I take issue with your first statement- evolution certainly works at the species level (in the sense that population genetics is a thing). It does appear there is a population fitness that is selected for when two desirable phenotypes can't be accomodated in a single individual. Also, "Most organisms die long before they get cancer, so the fitness of a gene that prevents cancer is pretty small." ignores the fact that there are genes preventing cancer (tumor suppressors), and those genes are under active selection.

rcthompson 10y ago
The term "tumor suppressor" doesn't necessarily mean that the gene exists for the sole purpose of preventing cancer. It just means that the protein's normal function happens to prevent cells from becoming cancerous. Many DNA-repair proteins fall into this category, for example, simply because repairing errors in a cell's DNA reduces the probability of that cell accumulating cancer-causing mutations. And yet, DNA-repair proteins pre-date the evolution of multicellular organisms, so they cannot possibly have evolved as a response to cancer. In practice, the term "tumor suppressor" generally means a gene for which recessive loss of function mutations are associated with cancer (i.e. you have to lose function of both copies of the gene in the same cell before there's an effect), while "oncogene" means a gene for which dominant gain of function mutations are associated with cancer (i.e. mutation in one copy of the gene is sufficient to have an effect). Neither term implies anything about how the gene evolved or whether selection due to cancer was a factor in the evolution of the gene.
- dekhn 10y ago
  While most of what you say is generally consistent with mainstream science (including my copy of Cancer by Weinberg, AKA, "the bible"), it neither refutes my point, nor provides any support for the parent poster's claim. Also, "DNA-repair proteins pre-date the evolution of multicellular organisms, so they cannot possibly have evolved as a response to cancer." is not a valid argument. There are eukaryotic DNA repair proteins not found in bacteria, for which evolutionary evidence shows they evolved more recently. A number of such genes aren't just DNA repair, but cell cycle controllers that respond to DNA damage and coordinate complex activity. Whether they are "sole purpose of preventing cancer", is a very deep question, and goes to the heart of the nature of cancer. For example, naked mole rats either never or very rarely get cancer; current thinking is that hyaluronase they produce has a protective effect. Did the genes that make enzymes that produce hyaluronase evolve "specifically to prevent cancer", or did they play other roles and just incidentally provide cancer protection? Those are challenging hypotheses to express and prove either way. I'm sure you're familiar with the Hallmarks of Cancer.
Houshalter 10y ago
>evolution certainly works at the species level (in the sense that population genetics is a thing). That's not at the species level. That's, at the very best, at the level of a very small group. And most of those theories have generally been discredited - it only works if the beneficial effect on the group is sufficiently large, see http://lesswrong.com/lw/kw/the_tragedy_of_group_selectionism/ http://lesswrong.com/lw/kw/the_tragedy_of_group_selectionism... >there are genes preventing cancer (tumor suppressors), and those genes are under active selection. That doesn't contradict what I said. Those genes work well enough to prevent cancer in young organisms, but they are clearly not enough to stop all cancer. Why hasn't evolution evolved away all cancer? Because even if there was a gene that could decrease cancer risk by an additional 1%, it wouldn't actually increase fitness that much. A 1% decrease in risk is small on it's own, and then it only affects the 1% of organisms that haven't already died of other things. Preventing cancer is just not in evolution's priorities - at least not past a certain point.
- dekhn 10y ago
  I concede I was probably inaccurate in my statement about evolution and the species level; note, however, that I disagree with people who believe that the gene is the unit of evolutionary selection (I would refine it to say "functional region" rather than gene because I believe that mutations in enhancers and other transcriptional activators play a bigger role in evolution that mainstream science). On to your second reply: I basically agree. We're mostly arguing about where that certain point is. Problem is: cancer is a disease that is closely tied to many necessary functional parts of multicellular organisms. In a sense cancer is what gets people who weren't got by something more acute earlier, and probably evolution is addressing the more acute issues with a higher priority.

untilHellbanned 10y ago

I think you are arguing on two different levels. I'm a professor biologist and though it strikes me as a pretty indirect mechanism (arguing the effect size could be small), I see some merit to the commenter's idea.

rcthompson 10y ago
The current hypothesis is that humans have a longer lifespan because having living grandparents conferred a survival advantage over having only parents. Arguing that elderly relatives confer a disadvantage would seem to directly contradict that hypothesis, unless somehow grandparents were beneficial but great grandparents were not.
- untilHellbanned 10y ago
  A current hypothesis. There is always room for more ideas.

wyldfire 10y ago

> Because evolution doesn't work on a species level, it works on a gene level. Genes that help propagate themselves spread more. I don't think that's necessarily the case. If my descendants were able to reproduce more because of my gene that reaped me before I was a drain on their resources, that gene could be inherited by them. IIRC I vaguely recall something about male homosexuality being correlated with female siblings' increased offspring count. It's a stretch to say that they'd be caused by the same gene but it would be an interesting phenomenon which might be a net benefit to the species.