Scientists from the Moscow Institute of Physics and Technology (MIPT) developed a highly accurate and relatively fast analysis method that can predict how cell membranes will respond to the molecules of drugs and toxins. This will make it possible to calculate the effect that a medicine will have on the cells before any experiments. The research has been published in the Journal of Chemical Theory and Computation.
“The distinctive feature of our method is that it provides us with a complete description of changes in a molecule. We can track positions of all the atoms at once, assigning specific values to every structure variant, which we can use later for statistical analysis,” says the corresponding author of the study and the head of the Laboratory of Structural Analysis and Engineering of Membrane Systems at MIPT’s Center of Molecular Mechanisms of Aging and Age-Related Diseases, Ivan Gushchin.
Simulating the behavior of biomolecules is highly problematical because the movements of every atom have to be described – even a small molecule of 54 atoms requires 156 sets of values. The giant amounts of information acquired during the simulation are hard to process and interpret, and it takes a lot of time.
The research group led by Gushchin found a way to considerably simplify the analysis of the simulation results without a significant decrease in accuracy. To process incoming information, scientists used principal component analysis (PCA) – a method that selects the most essential data. It reduced the amount of data required for the research 10 times, with a decrease in accuracy of only 10%.
This approach was tested on an experimentally well-studied DOPC lipid (dioleoylphosphatidylcholine). For the simulation, scientists chose eight different force fields – parameter sets describing interaction between all the atoms. Some force fields are used for a very detailed description and others for a more approximate one. The researchers then applied the PCA.
It turned out that the motion description of a 54-atom molecule requires only 14 components – collective movements of a certain atom ensemble. For example, one of the components is responsible for two DOPC molecule hydrophobic tails moving apart in a scissor-like motion.
Two principal components of DOPC molecule motion. “Compact” and “extended” structures are represented by the red and blue lines on the picture. Transitional structures are in shades of purple. Image courtesy of the authors of the study.
Lipids are the building material for a cell membrane and in order to affect the cell every molecule has to pass through it. This is why studying lipids with single-atom accuracy will help predict the effect of drugs and toxins on cells and whole organisms using computer only, thus accelerating the search for new medicines and their testing. Simulation can also be useful in aging research, the mechanism of which is thought to be related to changes in the structure of molecules of cell membranes.
Structure of a DOPC molecule. On the left is the chemical structure. On the right are the conformations adopted by the molecule in molecular dynamics simulations. Image courtesy of the authors of the study.