The Fukui indices still tell the same story.
But the frontier has moved.

Enjoy this AI-generated image of Quek with (a somewhat creative representation of) Fukui indices, and read the blog at rush.cloud/blog/fukui-i... for the code to recreate the study yourself.

This isn’t just about scaling simulations; it’s about giving every researcher the means to turn ideas into impact.

Read more about Dr. Rojsitthisak’s approach at rush.cloud/blog/thailan....

Using Rush, his team runs deep-learning-enhanced virtual screens and ligand-guided docking without high-performance computing clusters or specialized coding skills. The result? A generation of chemists fluent in both chemistry and computation.

Two decades after studying drug-DNA interactions, he’s now helping modernize Thailand’s research ecosystem by embedding computational methods into the daily workflow of young scientists.

It’s quantum chemistry meets caffeine addiction. And yes, it ends with a pour-over recipe.

Now please enjoy this AI-generated image of Ads with her coffee, and then go read the post at rush.cloud/blog/geek-on....

In our latest blog, Adelene Sim dives beneath the surface to show how caffeine hijacks your brain’s sleep receptors, slips perfectly into the adenosine pocket, and even reveals which atoms do the heavy lifting.

Now please enjoy this AI-generated image of someone who doesn't really look like Yiyang at all, in what the AI gods have titled 'Medicinal Chemist Chemistry World' (look, our prompting may have been off for this one), and then go read the blog at rush.cloud/blog/computa...

If you’re still waiting on comp chem bandwidth, you’re already behind.

In our latest blog post, med chemist Yiyang See shows what it’s like to run a 1.7M compound virtual screen and a full docking workflow in natural language: no coding, no setup headaches. Just results on your desk by the next morning.

Still, the explosion of new computational platforms can feel overwhelming even to experts in the field; let alone to most medicinal chemists (myself included), who usually receive this information downstream."

It’s a reminder that even the most elegant simulations can mislead us if we don’t respect the complexity of conformations and environments.

Enjoy this AI-generated image of Ross climbing a free energy curve, and then read the details at rush.cloud/blog/inactiv....

- Why estimating free energy barriers is harder than it looks
- How Umbrella Sampling (US) works—and where it fails
- Why starting conformations and reaction coordinates can make or break your results
- A mental model for barrier heights that can guide better protocol design

In our latest blog, Ross Pure explores:

That’s exactly what virtual screening gives you.
And you can do it free, right now, on Rush: rush.cloud. Just type:

'screen for binders to pdb 9C14' (or your preferred target)

…and see what turns up.

Screen purchasable compound libraries or generate similar small molecules easily at rush.cloud.

Then, our similar small molecule generation uses fragment-based generation that builds from scaffolds with high synthetic accessibility scores, boosting the likelihood that novel compounds are also makeable.

🔹 Deep learning generators
Neural nets can bias toward activity, but they generally don’t account for synthesis - feasibility checks come too late.

At Rush we encourage users to start with reality by screening large purchasable libraries that cover chemical space and guarantee synthesizability.

🔹 Fragment-based approaches
Pieces molecules together from functional fragments. Strikes a better balance, but careless linking can still create lab-infeasible designs.

🔹 Atom-based approaches
Like sketching molecules atom by atom - high novelty and diversity, but many are impossible to make.

🔹 Reaction-based approaches
Built on known reactions, so you get compounds you can synthesize. But novelty suffers - lots of close analogues.