All-atom Molecular Dynamics Simulations of Meiosis 1-associated protein (M1AP) to Investagate the Impact of Known Missense Mutations Associated with Male Infertility through Non-obstructive Azoospermia

Protein structure of meiosis 1-associated protein (M1AP) was modelled by using two of the best tools in CASP13, namely Robetta (from Baker Lab) and GalaxyWeb (from Seok Lab). Even though some other available best CASP tools were also utilized for modeling, they were not included this dataset because of low quality (The related results are included the paper that will be published soon). After these two best models were compared by analyzing the stability via molecular dynamics (MD) simulations, GalaxyWeb model were used to investigate the impact of mutations (i.e., S50P, R266Q, P389L, G317R, and L430P) in M1AP which were recently associated with non-obstructive azoospermia (NOA). NOA is a male infertility-related condition causing absence of sperm in the seminal fluid due to meiosis failure. We aimed to elucidate the pathogenicity mechanisms of these five missense NOA-related mutations on M1AP by performing molecular modeling and MD simulations. This dataset includes the results of 500 ns MD simulations (two repeats, each 250 ns) for each of the mutant and wild-type systems. Systems were prepared in Visual Molecular Dynamics (VMD 1.9.3) by placing them in a TIP3P water box with approximately 20 Å thickness from the protein surface and neutralizing the system charge with 0.15 M KCl. Of note, only protein parts were kept for the submission to reduce the size of files. Nanoscale Molecular Dynamics (NAMD 2.13-CUDA) was used to perform MD simulations with CHARMM36m force field. For pressure and temperature controls, Nosé-Hoover Langevin barostat and Langevin thermostat were used. ShakeH algorithm of NAMD was applied for water molecule constraints. 12 Å cut-off distance was used for van der Waals interactions. Switching function starts at 10 Å and reaches zero at 14 Å. Integration time-step was 2 fs. To compute the long-range Coulomb interactions, the particle-mash Ewald method was used. NPT ensemble was applied for whole simulations. Two step minimization & equilibration procedure was performed: (1) 5,000-step minimization and 1 ns equilibrium with constraints on the protein; (2) 5,000-step minimization and 1 ns equilibrium without the constraints on the protein. All related configuration files for wild-type system were also included to the dataset. Production simulations were run twice along 250 ns by using different random seeds to assign the velocities from Boltzmann distribution (total simulation time for each system was 500 ns, which are given as 250 ns repeat 1, and 250 ns repeat 2). Only the production simulations were supplied in the dataset whereas the minimization and equilibration trajectories are available upon a reasonable request.