The 'show why science is awesome' thread:


#983

This is something I’ve been following for a couple months now and is starting to get some widespread attention:

Basically, you take some white blood cells (called ‘T-cells’) from a patient’s body, modify their genes using a fragment of the HIV virus (!) to make them cancer-killing machines, and then re-insert them into the patient’s body. The T-cells then multiply and attack the cancer in a way that the un-altered T-cells would not.

The “early-days” results against leukemia patients have been astonishing, with an 83% remission rate (and with 66% remaining cancer-free).

The down-side is that the therapy is necessarily bespoke. It’s never going to be a pill you can stock in a pharmacy. And although the $ half-million per-person price tag will surely come down, it’ll probably never be a cheap process.


#984

One of the crazy things about parts of genomics/genetics is the price scaling. The cost to sequence the whole human genome back in the late 90’s was ~ 2.7 billion dollars. We can sequence a person’s whole genome in a day now for about $1,000. We routinely run this level of sequencing in the lab as “no big deal.”

I’m pretty sure as gene therapy gets safer (more accurate CRISPER-like technology) and we have a better survey of exactly what changes are effective given someone’s genetic background, this will move from a specialized lab process to something that can be cranked out by well trained lab techs and software. I’m not sure how long that’ll take, but “never” is a really long time.

Some labs (like the one I’m in) are always thinking about how to scale our studies/analysis, and most of our tech dev work takes processes that cost thousands of dollars per and get it to a few dollars by inventing new technology/analysis to get around the bottlenecks.

Obligatory data:


#985

Yeah, using “never” when talking about a technical hurdle is a rookie mistake that I should have known better than do. “It will be a while before the cost of this comes down to the point where it’s available to the masses” is probably more accurate.


#986

AI Obama

[quote]
The researchers were careful to not generate videos where they put words in Obama’s mouth that he did not at some other time utter himself. However, such fake videos are “likely possible soon,” says study lead author Supasorn Suwajanakorn, a computer scientist at the University of Washington.[/quote]


#987

Still a better president than He Who Will Not Be Named.


#988

I’ve estimated that the lifetime costs of my son’s leukemia treatments and subsequent bone marrow transplant will be in the $2.5 million+ range.


#989

Holy crap! That certainly adds a new perspective.


#990

I for one am ready for our Raven overlords.

He described to me how one experiment took an eerie turn: One raven in the experiment figured out how to work their rock/box contraption first, then began teaching the method to other ravens, and finally invented its own way of doing it. Instead of dropping a rock to release a treat, the future Ruler of the Raven Kingdom constructed a layer of twigs in the tube, and pushed another stick down through the layer to force it open. The bird had to be removed from the experiment before it could teach any other birds how to do it.


#991

Sorry, Raven. Nevermore.


#992

We have laser weapons.


#993

I imagine if the second gen can take down missiles this will be far more accurate than the other interceptor-type weapons systems the US has been trying to develop as a missiles shield.


#994

Here’s a way that science is not so awesome. Or more accurately, why we need to really careful about what we believe in science journals, because their review methods may not be so awesome.

Accepted by four journals, despite a line saying “The majority of the text in the current paper was Rogeted from Wikipedia” in the paper itself.


#995

The problem being highlighted in that blog post is the ‘predatory journals’. Science publication used to be strictly the realm of paper journals where the cost of publication was borne primarily by subscription fees (mostly paid by university libraries, but many journals also had a significant number of individual subscribers). A rebellion of sorts began in the 90s, with the confluence of certain publishers making an exorbitant amount of money by jacking up subscription prices (Elsevier) plus the internet - with a few visionary folks realizing that paper was no longer a necessary part of publication. They led a push for ‘open access’ publishing, with cost of keeping the enterprise running being borne by a charge paid by the authors, rather than the readers. The unintended consequence being that pretty much anyone who has ever published a scientific paper now gets 5-10 emails per week from online ‘journals’ that exist only to extract publication fees from authors. They talk big about peer review, but that’s just a cover story - they’ll take anything from anyone who is willing to pay them money. All it takes to be an ‘academic publisher’ in that sense is an email address and some imagination to dream up sciency-sounding journal names.


#996

This is super cool.


#997

Amazingly well preserved ankylosaur goes on display


#998

Life that naturally thrives in pure acid. They are called polyextremephiles since they survive a triumverate of challenges. They’ve adapted to extreme acidity, very high temperatures, and high salinity all at the same time. In retrospect, it seems like life living in the liquid zones of Europa and Enceladus have fewer potential extremes to overcome than these critters.


#999

I wasn’t sure if I should put this here, or in the Grimoire thread, in honor of Cleve. We share our lab space with the speaker (David Reich), and he’s both a great statistical thinker, as well as very interesting guy.


#1000

There are a million of these stories all the time but this one seems more promising since this tech is already in use in some places they’re just making it better suited for general purposes:

Also I thought we had a battery technology thread or something similar but hopefully this will do.


#1001

#1002

What could possibly go wrong with allowing these things out into the wild?