The kinds of metrics they are urging for generative model papers are ones the comp chem community has been discussing for years: full disclosure of training sets and similar molecules, comparison with simpler techniques. A feeling of deja vu many times over.
-
-
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
And the hype from the investors pointed out ("Pharma's AlphaGo moment") is just absurd and indicates an abdication of responsibility. There needs to be a co-evolving circle where scientists and investors police each other, otherwise the hype will get completely out of hand.
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Ultimately I think one of the most promising things will be an automated journal submission process that simply doesn't allow papers to be submitted unless there's accompanying full data and code. The community needs to work on these kinds of peer-review validation tools.
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Too bad this is behind a paywall
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
“The critique of Murcko and Walters...fails to recognize that...our goal was to provide the first demonstration of the effectiveness of a novel generative approach; as such, in-depth validation of the molecules produced was not the main goal of our paper.”https://www.nature.com/articles/s41587-020-0418-2?error=cookies_not_supported&code=cfb45ac8-a595-4328-bd8a-b6577b320095 …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.