Car drivers in Berlin: The first one ignores my green light and simply turns in before me. The second one screams at me that my light is now red (even though it was green when I started walking).
Janine George
@molecularxtal.bsky.social
Computational Chemist. 🦀🐍 Scientist at BAM + Professor FSU Jena. Only private opinions here. Last name: German pronunciation. 🏳️🌈 she/they (sie). @JaGeo on github
New preprint 🚀 : How can features learned by pre-trained machine learning interatomic potentials (MLIPs) be used for both materials generation and the evaluation of generative models? arxiv.org/abs/2607.287... #CompChem
Representations from Pretrained Machine-Learning Interatomic Potentials as Coarse Coordinates for Material Generation and Evaluation
Generative machine learning is increasingly used for inorganic crystal structure generation. Most models and the corresponding evaluation approaches rely on simple forms of crystal structure represent...
arxiv.org
A new preprint comes out and I am excited. Weird, actually, after all those years. But I guess I would not be able to work like this if I wasn't excited!
My longest paper project will hopefully soon come to an end. 😅 I have asked about this paper for the first time in 2018!
The 1951 Festival Pattern Group transformed X-ray crystallography—including Dorothy Hodgkin’s insulin diagrams—into modern manufacturing designs, blending post-war science and art into innovative national progress.
In heart, I am a chemist. I think in graphics rather than equations.
We have been collecting contact infos for many, many coauthors. 😬 Still, a few people cannot even be identified...
Spannend ist wirklich, dass Spahns Rücktritt wirklich auf der Straße diskutiert wird. Gestern im Restaurant, heute im Bus. Manchmal dann noch inklusive größerer (homphober) Monologe rund um das Thema. Aber zumindest nicht unwidersprochen.
So, on LinkedIn, I did this post on some challenges I had up to my current position. I am aware many people had them. This does not make me more worthy or anything. I just had a really bad day yesterday and had to remind me that it can be worse and most of the time does get better.
Lot's of wins today but also some losses. Not even fully motivated to really celebrate my old/new job.
I am watching two soaps (UK/Germany). On both soaps, the murderer of recent episodes confessed the murder to their closest person exactly today (in TV, the German one is one week behind) It's a bit boring, tbh.
I think I have to get rid of Instagram again. The mindless scrolling wastes so much time. I could actually read instead....
Great to see people using agents to write new interfaces to other codes for which nearly perfect solutions already exist. But now rewritten with an agent, partially reusing the packages and without test code. 😅
Wenn ich durch die 200 Leute manuell durchgegangen bin, verdiene ich wahrlich einen Orden. Wtf
Okay, my body refused to wake up one hour ealier than usual. 🫠 Now, I had to rush.
Habe 4 h geschlafen, mein Zug hat 1 h Verspätung, der 30 min davor ist nicht aufgetaucht. Hier im Abteil piepst es schrecklich und ständig läuft jemand an mir vorbei auf die Toilette, weil der Zug zu voll ist. Bin super gestresst.
One of the last lecture days in this semester today and my brain is in pieces again. Will try to patch it together with coffeine. I am awake since 5:30 and I wasted most of the time staring on weird IG posts. Hello.
It's 13:00 and we just hit 39°C in Berlin. We are insanely privileged to live in the green part of Berlin and to be able to afford a mobile AC. I cannot wait for the weather change.
I am planning a vacation to visit my fav soap sets. 😅 The other short vacation will include roller coasters. (People are usually surprised that I like roller coasters.)
Amazing postdoc in my team Phil Harrison just preprinted on @chemrxiv.org his extensive review on biologically-conditioned molecular generative models, where we also provide perspective on where gaps are in the field and the challenges spanning biochem domains #chemsky chemrxiv.org/doi/10.26434...
Biologically-driven generative chemistry: using biological data to guide de novo drug design | ChemRxiv
Drug discovery is a dauntingly difficult process due to the enormous space of possible molecules and the complexity of biological systems. Traditional in silico drug design methods have focused primarily on optimising chemical properties, often ...
chemrxiv.org
I still find it fascinating that people suddenly start listening to what you have to say simply because of a certain position, qualification, or general fame.
I was shortly thinking to submit a contributed talk to a conference in September. Then, I got 4 (?) invitations within two weeks for this and next year and I decided that my August and September will be talk-free. 😅♥️
Last weeks hotel experience: "Good luck to find something glutenfree" This week: "We make everything freshly in the kitchen for you to avoid crosscontamination of the buffet" 😅
They make the joke that having to live with a glutenfree diet is a good cure for social anxiety. 🫠
Also: heat wave in May. 🫠 The regional train is still soo hot.
I am still on time with my train as I am smarter than the DB. I started my regional train trip an hour early, they than opened up my ticket for all trains and I took an intercity train nearly 1 h early.
I never get: my papers are not good enough for certain high impact journals, yet they want me to judge whether other people's work is good enough for the same journal. Mh. Also: Currently reviewing a phd thesis instead ...
I am someone whose intelligence is often only discovered when one looks 3 times or when I had enough time to prepare, I am not tired and full with coffeine. Consequently, I have always doubted myself a bit too much.
I was at the exhibition "Ruin and Rausch, Berlin 1910-1930". Very small exhibition. While core topics were discussed, I missed the connection with history and science.