gpcr1949
Joined 6 years ago
Karma
297
Submissions
73
Member since
Sep 25
my main interests are chemistry and audio.
View on Hacker NewsRecent submissions
Comment · 1 year ago
I think it is an underestimation to say ~800 people died in the Gaza war. The Ukraine number also seems like a serious underestimation.
Comment · 2 years ago
Israel is committing what some experts are calling a genocide. In any case, they killed a lot of civilians. It can probably be debated whether this is noble or "striving to be better". But either way, the political beliefs and hobbies of the laureates aren't interesting to me, and I can't imagine the Nobel committee seriously taking this into account.
Comment · 3 years ago
I am really surprised they fired him for a really relatively mild series of tweets. I know some of the Israeli scientists leading the campaign against him also pointed out he tweeted "Fuck Israel" in 2018 [0]. Due to the date, May 15 2018, he is probably referring to IDF slayings the preceding day during Palestinian border protests [1]. Nature also did a factual write up of this guy's firing [2].
[0] https://twitter.com/mbeisen/status/996404294529138688
[1] https://en.wikipedia.org/wiki/2018%E2%80%932019_Gaza_border_...
[2] https://www.science.org/content/article/prominent-journal-ed...
Comment · 3 years ago
Writing Basel as Bâle would be silly and non-standard too, yes. Maybe "wrong" is too strong of a classification. But for example, I wouldn't immediately recognize it as Basel in a headline. FWIW Leuven refers to itself in English as Leuven on their own website: https://www.leuven.be/en
Also somewhat interesting to me is that for some reason people think Belgium is mainly a French speaking country, but in fact there are actually more Dutch speakers than French speakers (56% vs 38% native speakers according to Eurostat in 2015) - I have the sense referring to Flemish cities by their French name reflects that
Comment · 3 years ago
Interesting history, but people in Leuven call this city Leuven and not Louvain. No need to use the french term in a flemish city where almost no one speaks french as their first language. In fact, french language was the subject of some amount of political struggle in the 1960s, resulting in the founding of the french speaking university in "Louvain-la-Neuve", a planned city that was built in (french speaking) Wallonia for this purpose. You can read more about the politically important language discussion that occured in the 1960s here: https://en.wikipedia.org/wiki/Split_of_the_Catholic_Universi...
Comment · 3 years ago
Comment · 3 years ago
They use a random forest classifier, which is an ensemble model that gives a consensus result of several decision trees. One way to achieve this consensus is voting. Random forest models are commonly used in building chemical models like this (and in QSAR), because they are quite robust. Due to the typically small size of chemical data sets (dozens to thousands, typically), more sophisticated methods are not usable and do not perform better.
Comment · 3 years ago
High-throughput screening (HTS), i.e. testing a large amount of compounds in simple biochemical assays can't be compared with testing 10 compounds in monkeys. The point of HTS is finding active substances in a large collection of compounds (finding needles in the haystack), and then using this as the initial step of a whole chemical optimization research program that may lead to a new drug candidate down the line.
Conversely, the 10 compounds tested in monkeys will usually be advanced drug candidates just before phase 1 trials in humans. If you would give monkeys 10 compounds from a HTS collection you would learn nothing about what works in human disease because most compounds are inactive at a given target.
Every large pharma company has HTS infrastructure and uses this for early discovery (in addition to animal experiments at later stages of drug discovery campaign) - and these companies are interested in making profit by discovering new drugs that meet the requirements to get approved, not in generating publications or boasting about their fancy HTS robot.
Comment · 3 years ago
You can read about the development of this new route and its advantages here [0]. Ctrl+F for ibuprofen or go to p.54.
Comment · 3 years ago
Agreed the process is very simple. The innovation that makes the resolution worth it is (as Derek also mentioned in the blog post) an efficient way to re-racemize the L-meth to DL-meth, which can then be resolved again. With steady state production and pooling this means it can be done using just one batch size (and not ever smaller "brsm" style bathces). If these steps are efficient enough you definitely get much more bang for your buck. Here is one such method[0] which probably inspired these drug manufacturers (as they use thioglycol and AIBN, which has been found in clandestine lab seizures).
[0] https://sci-hub.se/10.1021/jo061033l (published in JOC in 2006 - so only a mere 2 decades behind the curve)
Comment · 3 years ago
This is really a scandalously misinformed stunt. The chemical they released is SO_2 - one of the gasses and car pollutants causing acid rain.