After going through the libraries mentioned in that post, it seems like BioStructures.jl and Chemfiles.jl might not be as robust as MDAnalysis in Python. Additionally, it appears that MDToolbox.jl may have ceased development.
I’m eager to know if there are any newly developed and comprehensive libraries that have emerged in recent years. I’ve noticed that MDAnalysis can be very slow when dealing with large trajectories. Any insights or recommendations on faster alternatives would be greatly appreciated. For example, the distance related analysis.
I don’t know of any more beyond those mentioned on the linked thread. The Julia ecosystem is lagging a bit in this area.
I’m not sure how much faster MDAnalysis could be. If there is performance left on the table then maybe someone will be motivated to make a new package. If MDAnalysis is close to optimal then it’s a lot of work to get to something with the same functionality.
I didn’t do a benchmark for MDAnalysis. The concept of it’s slow comes from intuiation and practice. When performing the same distance analysis between an atom pair. The Cpp based Gromacs command gmx distance is much faster than do the same thing by MDAnalysis. Because I was dealing with a trajectory with 10^6 frames. I just admire the comprehensive functionality and ease of use of MDAnalysis.
I think a Julia version MDAnalysis will be faster because when it comes to a lot of calculations, Julia always wins in speed.