TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Modeling Molecules with Recurrent Neural Networks

35 pointsby csvossover 9 years ago

3 comments

gene-hover 9 years ago
That is a very strange way to generate molecules. It would be interesting to see if this is any better than other approaches used to generate molecules like those used in drug design. It would be particularly interesting to see how this compares with the approach that was used to generate GDB-17[1] a database of randomly generated molecules or at least see if generated molecules pass through the filters used to make GDB-17. Grammatically correct does not necessarily mean physically reasonable, recall Chomsky&#x27;s &quot;colorless green ideas sleep furiously.&quot;<p>While it would be interesting to model reactions with RNN, I&#x27;m not so sure this would offer any advantage over simply searching a database of reactions like the Crossfire Beilstein database[2][3]. I am also curious if you investigated reaction MQL in work with the carbonate project.<p>This work is interesting though. Essentially you are using a neural network to generate graphs. There are a lot of things that can be represented with graphs, IE electric circuits and such. Maybe you could make a RNN for generating neural networks!<p>[1]<a href="http:&#x2F;&#x2F;www.gdb.unibe.ch&#x2F;gdb&#x2F;home.html" rel="nofollow">http:&#x2F;&#x2F;www.gdb.unibe.ch&#x2F;gdb&#x2F;home.html</a> [2]<a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Beilstein_database" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Beilstein_database</a> [3]<a href="http:&#x2F;&#x2F;www.ncbi.nlm.nih.gov&#x2F;pubmed&#x2F;21378798" rel="nofollow">http:&#x2F;&#x2F;www.ncbi.nlm.nih.gov&#x2F;pubmed&#x2F;21378798</a>
Xcelerateover 9 years ago
Haha, interesting, but I thought this was going in a different direction. There&#x27;s a lot of work going on lately to try to model molecular potentials (for molecular dynamics &#x2F; quantum chemistry) using neural networks. Accurate potentials are extremely expensive to calculate.
评论 #10383638 未加载
bloafover 9 years ago
Where did your data set come from? It contains at least one error:<p>WOF2, tungsten(VI) oxytetrafluoride<p>that correct formula is WOF4
评论 #10384502 未加载