Mdlvzip (RELIABLE — 2024)

: Despite high compression, it preserves ensemble-level features like RMSD fluctuations and radius of gyration.

The core of MDZip is a neural network called a . Mdlvzip

: The decoder reconstructs the atomic positions with high geometric accuracy. : Despite high compression

: It shrinks data by more than 95% compared to raw formats. Mdlvzip

: You can share compact "latent representations" instead of entire raw trajectories, making collaboration easier.

: Users can select different bottleneck sizes to balance how much space they save versus how accurate the reconstruction is.

: The encoder takes the complex atomic positions and maps them into a tiny "bottleneck" size.