Hacker News Reader

xyhopguyon American Academy of Sleep Medicine calls for elimination of daylight saving time

waking up two hours befor sunrise is miserable no matter how dark it stays. In anycase, for anyone in the northern US the sun would still set before or at 5:30.

xyhopguyon American Academy of Sleep Medicine calls for elimination of daylight saving time

in seattle that means sunrise at 9 and sunset at 515 in a place that's already hella dark. No thank you.

xyhopguyon Portland passes low-density zoning reform

in the denser parts of portland it already is!

xyhopguyon Portland passes low-density zoning reform

densify portland -> improve tax revenue -> improve transit. nyc certainly benefited from aristocrats and lax labor laws when building the subway, but this isnt something that is impossible for portland. trimet is already huge for a city as small as portland. improving density will only help transit services.

it doesnt have to be subways -- the willamette flood plain isnt great for subways anyway. portland should continue to build out its bus service and cycling infrastructure. portland is already pretty reasonable for living without a car, it will only improve with density.

xyhopguyon Fast high-level programming languages

Problems like the benchmark in the post are actually really simple to write in rust. They're all essentially streaming algorithms with some simple parsing.

Fwiw rust has completely replaces cpp for high performant bioinfx code in my workflows. Sooo much easier to write than cpp!

xyhopguyon New Abbott SARS-CoV-2 antibody test has 99.90% specificity and 100% sensitivity

whoops, I presumed you were asking about the math. My source is nytimes which pulls from the dataset you linked.

https://www1.nyc.gov/assets/doh/downloads/pdf/imm/covid-19-d...

Specifically has numbers for NYC. You can sum the boroughs or use the total listed below. Using confirmed by pcr cases you get ~15k/8.6 = ~.175% as you said.

Using probable excess deaths adds an extra 5k deaths, hence ~.25%

xyhopguyon New Abbott SARS-CoV-2 antibody test has 99.90% specificity and 100% sensitivity

they have a different lineage, but as I stated previously, no major functional mutations were observed.

xyhopguyon New Abbott SARS-CoV-2 antibody test has 99.90% specificity and 100% sensitivity

That's a great point although obviously its very hard to estimate something that has a horizon of atleast a month. the bay area did shut down quite a bit earlier, it could be that the lack of deaths is simply due to a better response.

xyhopguyon New Abbott SARS-CoV-2 antibody test has 99.90% specificity and 100% sensitivity

hm, care to explain more? that's exactly how I think about it. Total deaths in NYC attributable to covid (~20k) over the total population (8.5M) = ~.25%. In my mind this is a very easy lower bound to estimate, what am I missing here?

In anycase, even a rate of ~.1% would be absurd given ~20% of people are testing positive by antibody tests.

xyhopguyon New Abbott SARS-CoV-2 antibody test has 99.90% specificity and 100% sensitivity

afaik there havent we havent observed major functional mutations in the virus

xyhopguyon New Abbott SARS-CoV-2 antibody test has 99.90% specificity and 100% sensitivity

people say that because .25% of the population in NYC are dead from covid. we dont see those numbers here -- if it was widespread in feb, what caused the discrepancy?

xyhopguyon Researchers: Are we on the cusp of an ‘AI winter’?

https://en.wikipedia.org/wiki/Odds_ratio

odds of crash a => 97% => 3 / 100

odds of crash for b => 99.99% => 1 / 10000

improvement (odds ratio in this case) is then 300x = odds of crash for b / odds of crash for a

xyhopguyon 'Cancel Culture' Comes to Science?

all genes on X is a meaningless statement unless all genes are involved in X... if you read the paper the authors seemingly put quite a bit of effort into the rigor of their model.

they used multiple validation cohorts, they controlled for environmental factors, they controlled for non-autosomal factors, applied multiple methods to boost the predictive performance of their model. adding more snps typically leads to an overfit and poor performing model on validation cohorts, which is why they use so much rigor and end up with usually 100-400 snps for each of the phenotypes.

in anycase they found a lot more than 10 SNPs, and they did more than one analysis in the paper.

Parent educational status is a significantly better predictor than the polygenic risk score for educational attainment. heritability (also known as r^2) for cognitive performance they estimated at an even lower ~7-10%.

xyhopguyon 'Cancel Culture' Comes to Science?

https://www.nature.com/articles/s41588-018-0147-3

xyhopguyon Genomics – A programmer’s guide

have you ever looked at the test suites for Picard? All regression tests and the library is OO hell lols

I was taught a decade ago that rolling your own in genomics isn't as bad of a decision as it seems.

xyhopguyon The Heart of a Swimmer vs. the Heart of a Runner

I dont think this is settled -- I've seen world class swimmers do it both ways. Personally I find taking a long stroke after each breath more effective.

xyhopguyon The rise and fall of scientific authority, and how to bring it back

is trusting scientific authority the same thing really? Believing the theory of evolution is different than trusting that the $100 genome is around the corner... And unlike religious institutions, scientific institutions typically reflect that difference in their rhetoric.

xyhopguyon One year into legal pot sales California doesn’t have expected bustling industry

California's neighbors to the north both have 1500+ establishments. Oregon isn't exactly big, but due to massive overproduction and no sales tax prices in Oregon are approaching sister crops like hop.

Wtf is going on in California.

xyhopguyon One year into legal pot sales California doesn’t have expected bustling industry

Didn't stop Oregon. We've got 5x as many dispensaries as California and wayyyyyyyy less people. Oh and the prices are like 75% lower.

xyhopguyon Driverless Hype Collides with Merciless Reality

The tough part about this line of reasoning is you can't tease apart the discriminatory power of human knowledge from human vision. How good does our expert system need to be? How good do our sensors need to be? I'm not so sure this is an easy to question to resolve

xyhopguyon Dads Pass on More Than Genetics in Their Sperm

I personally believe it's all integrated together. Sure methylation alone isnt going to effect the genotype of an individual, but thats not day it won't trigger some translocation event, expose LTRs, or some other mechanism of changing the genome. In everything I've seen thus far, I'd guess that it's all a complicated and interwined mess. How do you distinguish evolution and selective pressures of an individual from populations anyway?

xyhopguyon IBM's Watson recommended 'unsafe and incorrect' cancer treatments

Here's a nice review: https://www.ncbi.nlm.nih.gov/m/pubmed/28233803/

a biopsy from your peripheral plasma, opposed to solid tumor. A big issue in cancer medicine is it's super fucking hard to get any kind of noninvasive measurement. Typically it's done with surgery or with CAT scans, which have extremely low precision.

Research in the last few years has been pointing to the idea that we can detect tumor derived DNA fragments in plasma. The challenge being, healthy dna makes up 99.9% of it, this means that currently the methods only work when the tumor burden (metastatic for instance) is high. Not ideal for early detection or treatment monitoring. But if the computational tools improve (rn none assume such a low mixture), you could see sensitivity and precision increase to the point that it's useful for predicting therapeutic outcomes and for early detection.

xyhopguyon IBM's Watson recommended 'unsafe and incorrect' cancer treatments

I work in this space. The breakthrough will most likely come from computational biologists, not an external ML group. Liquid biopsy is close and that's the domain that reeeallly needs ML/AI in additional to strong mechanistic models.

xyhopguyon IBM's Watson recommended 'unsafe and incorrect' cancer treatments

Not if you understand mechanism. In reality the breakthrough will be biologists applying ML/AI, not the other way around. Domain and mechanistic knowledge are king.

xyhopguyon Dogs rush to help when owners cry, study suggests

> you should never allow other people to feed or pet it.

Let children and elderly pet/feed and you suddenly have a well behaved protection dog.

xyhopguyon In 5 states, richest Americans live in a “new Gilded Age”

Not really what im getting at. The pnw was developed later so I'd expect it to have less old money -- look at Oregon. The tech industry has certainly done a lot for the upper middle class in Seattle but the article is talking about people making 400k+ a year..

xyhopguyon In 5 states, richest Americans live in a “new Gilded Age”

interesting that Seattle isn't anywhere to be found. I've always "felt" that the PNW had less old money. Anyone wanna feed my confirmation bias some more?

xyhopguyon Layoffs at Watson Health Reveal IBM’s Problem with AI

IMO AI/ML scientists are drastically underestimating the complexity of biology.

Once scientists in biology and healthcare get on board like they are in linguistics and computer vision I'd expect things to pick up.