robwwilliams

63 ms·

931.
by robwwilliams 5y ago
Lol. I usually read HN to follow tech trends. This thread in hilariously uninformed. It is as shaky as me writing a critique of database engines would be. On this specific subthread: yes the individual is the level of selection (mainly) but
932.
by robwwilliams 5y ago
Grishka needs to read more widely on this topic. Start with Caleb E. Finch’s magisterial book “Longevity, Senescence, and the Genome” (University of Chicago Press, 1990). Or read my review with Michael Hook: “Genetic Cartography of Longevit
933.
by robwwilliams 5y ago
The current best treatment for death is to contribute to social health and knowledge. Transmit both to your progeny and younger colleagues. Your particular atoms and molecules are really not all the unique and wonderful.
934.
by robwwilliams 5y ago
We would all be wise to plan along these lines until your 99% sure-thing is proved. I study the genetics of longevity and alas I am gracefully planning to die.
935.
by robwwilliams 5y ago
Yes, and genomes only a small but significant part of the longevity puzzle.
936.
by robwwilliams 5y ago
You mean like everything governments know about smoking, drug use disorders, poor diet, exposure to environmental toxicants, and how they have used that information with marvelous efficiency? Yep.
937.
by robwwilliams 5y ago
No, they will be smart enough to know this is bogus.
938.
by robwwilliams 5y ago
I have not read the original paper yet, but I can tell you that longevity is not controlled by any one process. Its complicated. Simple stories are simply wrong. That complexity is amplified by environmental factors and intense gene-by-env
939.
by robwwilliams 5y ago
Not right. Get some hard core background in population genetics. If you want a great (but costly) text on this topic with real insight—, then read Lynch’s The Origins of Genome Architecture.
940.
by robwwilliams 5y ago
Nope: there is a huge difference between MHC gene recombination and mutation.
941.
by robwwilliams 5y ago
Do you have a reference? It would be hard to establish causality. Genetic drift extremely important in megafauna due to very small population size. Selection has to be exceedingly strong. Read Mike Lynch’s lovely paper in PNAS in 2007 (free
942.
by robwwilliams 6y ago
A “just-so” story; not science. So many factors could have and probably did contribute to CNS enlargement. But no way for this article to be more than a plausible speculation. Most work on “human exceptionalism” falls into the just-so story
943.
by robwwilliams 6y ago
The battle between frequentist statisticians and those advocating Bayesian approaches is quite old—-back at least to Wright. Pearl is not inventing a dichotomy but explaining why models are necessary to evaluate causality. Is this genuine p
944.
by robwwilliams 6y ago
Agree with this counterpoint. Sewall Wright for one (and his father) are given great credit. It is RA Fisher who comes in for well deserved flak for his infamous obstinance.
945.
by robwwilliams 6y ago
Great book on this topic: Judea Pearl’s Book of Why. He discusses causal models in great depth. And a highly amusing read. Mendelian randomization can sometimes “rescue” causality from large epidemiological studies. Pearl explains how.
946.
by robwwilliams 6y ago
Yes, travesty—but caused by intense commercial competition and what you could call “hub war”. The new LG CX TV+ wants to be the media hub; the new Denon audio-visual receiver wants to be the AVR+ hub for all sources and outputs. Then we hav
947.
by robwwilliams 6y ago
Perfect commentary—thank you, thank you. Movie theaters or a great market opening for a single LG-Apple-Denon device.
948.
by robwwilliams 6y ago
I have not read the book Science Fictions, but will comment that the core problem in academic science is the intense structural reward of telling “stories” rather than focusing on high quality data generation and robust causal models. A sna
949.
by robwwilliams 6y ago
The real winner (if hacking this homomorphic encryption is robust) will be so-called precision health care. Failure to crack the encoding means that human data will be much more sharable.
950.
by robwwilliams 6y ago
This comment on quantitative protein assays is only partly correct. It is definitely harder than quantitative assays of messenger RNA (aka “gene expression” and not as comprehensive. But there are now hundreds of quantitative proteomic stud
951.
by robwwilliams 7y ago
Not probable and highly likely to be confounded by secular and social cofactors. These dietary models and recommendations in humans are dumb “one size fits all”. We know from extensive animal research that genome-by-environmental effects ar
952.
Worst (joke) ideas: A single syllable version of the English Language(worstideas.wordpress.com)
1 points by robwwilliams 10y ago|0 comments
953.
Why I published in a predatory journal(the-scientist.com)
292 points by robwwilliams 10y ago|113 comments