false_positives.md

Removing false-positive datasets

Due to the indexation of zip files, we might have collected some false-positive datasets, i.e. datasets that contain zip files and matched our keywords but that eventually did not contain any molecular dynamics files (after looking into the zip files).

False-positive examples:

• Dataset Eigenvalue self-consistent GW quasiparticle energies... has been found because of the expression Molecular Dynamics trajectory in the description. But all zip files contains mostrly .log and .punch files, not strictly MD data files.
• Dataset Sampled ΔH/Δλ and ΔH data from ABFE... has the Molecular Dynamics keyword. However, zip files contains .xvf files only.
• Dataset Genetic Variability among Complete Human Respiratory... has probably been catched because of the words Molecular and Dynamics in the title but do not contain any MD data.
• Dataset Homology Modeling of Dopamine has been catched because of the expression Molecular Dynamics in the title and in the text. The dataset does not contain real MD files. Only PDB files.
• Dataset Computational Study of HIV gp120 as a Target... contains snapshots from MD and Docking simulations.

Special cases:

• In dataset CLN025__2RVD_3000ns_mutant, the archive CLN025__2RVD_3000ns_mutant.zip has no preview. Probably because the file is too large. We remove these kind of datasets anyway.