Tools from our lab you can try in the browser, without installing anything.
sincFold — RNA secondary structure prediction
End-to-end deep learning that predicts the secondary structure of an RNA sequence directly from the nucleotides, with no thermodynamic model and no alignment. Published in Briefings in Bioinformatics 25(4), 2024 · 10.1093/bib/bbae271 · source
Paste a sequence and press predict. If you need one to start with, this is a tRNA:
GUAGUCGUGGGCCGAGUGGGUAAGGGCGAUGUGACUAGAAAUCGAUUGGGGUUCUCCCCCGCGCAGGUUCGAAUCCUGCCGACUACGCCA
The demo runs on our institute's server. If the frame above is blank, use one of these:
open in a new tab | HuggingFace mirror | Colab notebook | run it locally
KinderNet — train a neural network with your webcam
An outreach app for primary and secondary school students. You show the webcam a few examples of two or more things, train a small network on the spot, and then test it. The point students take away is that nobody programmed the rules: the network learned them from the examples.
It runs entirely in the browser, so the images and the trained model never leave the device and are gone when the page is closed — which is what makes it usable in a classroom.
open KinderNet | institute page | source