Home › Glossary
Glossary
The terms used in this manual.
The terms below are used in the sense given here throughout this manual and in the reports.
| Term | Meaning |
|---|---|
| AM1-BCC | A method for partial atomic charges of small molecules: a semi-empirical AM1 calculation followed by bond charge corrections fitted to reproduce electrostatic potentials. The charge model of GAFF2 ligands. |
| Block-averaged standard error | The standard error of a mean computed from averages of consecutive blocks of frames, which accounts for the correlation between frames. When it does not reach a plateau, it is a lower bound, and the report says so. |
| Buried surface area | The surface of receptor and ligand that disappears on binding: SASA of the receptor plus SASA of the ligand minus SASA of the complex, from the same frame. |
| Cap (ACE/NME) | Neutral acetyl (ACE) and N-methylamide (NME) groups placed at the ends of a chain break, so that no artificial charged ends appear inside a protein. |
| Checkpoint | A file holding the exact state of a running MD stage, written every five minutes, from which the stage continues bit for bit after an interruption. |
| Compute capability | The generation number of a graphics card of the supported make. The GPU engine needs 5.0 or higher; see Performance. |
| Constraint | A bond length held fixed during the simulation. Bonds to hydrogen are constrained by default (LINCS for proteins and ligands, SETTLE for water), which allows a 2 fs time step. |
| Cosine content | A measure of how much a principal component resembles a random diffusion. Values near 1 mean the component describes drift rather than sampled motion. |
| Cut-off | The distance beyond which short-range interactions are not computed directly; 1.0 nm by default. Long-range electrostatics beyond it are handled by PME. |
| DSSP | The method of Kabsch and Sander for assigning secondary structure (helix, strand, turn) from the hydrogen-bond pattern of the backbone. |
| Equilibration | The short restrained runs after minimisation (NVT, then NPT) that bring the system to the target temperature and density before production. |
| Fingerprint | A checksum of the settings and inputs of a stage. A changed fingerprint makes the stage, and everything after it, run again. |
| Free-energy landscape | The probability of the sampled conformations projected on two principal components, expressed as G = -kT ln(P/Pmax). It describes the frames that were sampled; it is not a converged free energy. |
| Fusion insert | A foreign protein domain engineered into a protein to help crystallisation, such as T4 lysozyme in a receptor. Geinium recognises six common ones and offers to keep or remove them. |
| GAFF2 | The second general AMBER force field, for drug-like organic molecules; used for the ligand. |
| GB-OBC2 | A generalised Born model of the polar solvation energy (Onufriev, Bashford and Case, model II), used by MM-GBSA. |
| HMR (hydrogen-mass repartitioning) | Moving mass from heavy atoms to the hydrogens bonded to them (here by a factor of 3), which slows the fastest vibrations so that a 4 fs time step can be used. |
| Hydrophobic contact | A ligand carbon within 0.40 nm of a receptor carbon that is bonded only to carbon and hydrogen. |
| kT | The thermal energy, Boltzmann constant times temperature: about 2.58 kJ/mol at 310 K. Free-energy landscapes are given in units of kT. |
| LINCS | The algorithm that holds the constrained bond lengths of proteins and ligands. At 4 fs it runs with two iterations. |
| Minimisation | Steepest-descent relaxation of the built system to remove clashes before dynamics starts. |
| MM-GBSA | An end-point estimate of the binding energy: molecular-mechanics energy plus generalised Born and surface-area solvation, averaged over frames of the trajectory. It ranks ligands bound to the same receptor; it is not an absolute binding free energy. |
| MTS (multiple time stepping) | Computing the long-range non-bonded forces only every second step; an option of the production run that gains some speed. |
| NPT | Constant number of particles, pressure and temperature: the ensemble of the second equilibration stage and of production. |
| NVT | Constant number of particles, volume and temperature: the ensemble of the first equilibration stage. |
| Occupancy | The percentage of analysed frames in which an interaction is present. |
| PCA (principal component analysis) | A decomposition of the C-alpha motions into independent directions ordered by the variance they explain. |
| Penalty | A score attached to a ligand parameter that was taken from a similar chemical situation by analogy. Above 50 Geinium adds a note, above 150 a warning. |
| Periodic boundary conditions | The simulation box is repeated in every direction, so a molecule leaving one face re-enters at the opposite one. Trajectories are corrected (molecules made whole and centred) before analysis. |
| pi stacking | Two aromatic rings stacked face to face (planes within 30 degrees) or edge to face (planes at 60 degrees or more), with ring centres within 0.55 nm. |
| pi-cation | An aromatic ring and a positive charge within 0.60 nm, the charge above the ring face. |
| pKa | The pH at which a group is half protonated. Geinium uses model pKa values for protein residues and rules for ligand groups. |
| PME (particle mesh Ewald) | The method for long-range electrostatics in a periodic system, using a grid (0.12 nm spacing by default). |
| Production | The unrestrained simulation after equilibration from which all results are computed; 100 ns by default. |
| Project file | The JSON file that describes a job completely and reproduces it with mdsim pipeline. |
| Protonation state | Which titratable groups carry a proton, and therefore the charge of each residue and of the ligand at the chosen pH. |
| Replica | An independent simulation of the same system that differs only in its starting velocities (seed). Three or more replicas give a standard deviation between independent runs. |
| Result bundle (.gnm) | One file that holds a finished job, with thinned trajectories, and reopens it on any PC. |
| Results package | The folder <name>_results with the PDF report, every figure as an image file, the raw data and a manifest. |
| RMSD | Root-mean-square deviation of atom positions from a reference structure after a least-squares fit. |
| RMSF | Root-mean-square fluctuation of each atom or residue about its mean position. |
| Rg (radius of gyration) | The mass-weighted root-mean-square distance of the atoms from their centre of mass; a measure of compactness. |
| Salt bridge | Oppositely charged groups of ligand and receptor with their closest charged atoms within 0.40 nm. |
| SASA | Solvent-accessible surface area: the surface traced by the centre of a probe sphere of 0.14 nm radius rolled over the molecule. |
| Scientific check | A rule the export applies to a job before writing the report. A blocking finding stops the report until it is resolved or acknowledged with a reason. |
| Seed | The number that starts a random sequence, for example of initial velocities. Different seeds make replicas independent; equal seeds reproduce a run. |
| Statistics window | The part of the production run over which table statistics are computed: by default all but the first 10 %, which are treated as equilibration. |
| Water bridge | One water molecule hydrogen-bonded to both a ligand atom and a receptor atom in the same frame. |