Halo Experts,
Hope you are doing fine.
I have conducted MD simulations to calculate permeability values of one polar and one non polar drug molecule through a Blood brain barrier mimic lipid bilayer. Both of my values are 150 times smaller then the experimental value. Can I publish it? LIke even in literature permeability values through skin etc biophysical model are overestimated from MD studies then the experimental values by many times. Please guide me out.
Thank You
There is no answer to this. You should be able to defend your results and justify them the best you can. At the end of the day, your name is going to be on the paper, so it’s your responsibility to understand the “publishability” of a result, its implications, and eventually its shortcomings.
Why are the results like this? Do you see any problem related to the charge of the molecules? Is the force field you used known to be good/bad for these results? What method did you use to predict the permeability? As it will most likely be based on free energy computation, did you check that the free energy is properly converged? Are you sure the simulations are long enough and you captured all the degrees of freedom of the solutes? As you can probably derive from the same free energy also the partition coefficient in lipids, how does this compare with the experimental partition coefficient of the solute (if available, otherwise with the octanol/water one)? Is the composition of the membrane similar to the in vivo one? Are there previously published results? What do they obtain? The questions are all part of the definition and proper exploration of a research questions, and you should be able to address them, at least to an extent, to discern whether or not the results can be trusted or not. This is also valid if the comparison with the experiment is negative, as having a proper computational results that doesn’t match the experimental value is still valuable per se in answering the question ‘can we use these (force field, solutes, techniques, membrane composition, etc.) to predict properly this quantity? Well, most likely not’.
Consider also that permeability results are hard to compare with computational ones, as the macroscopic measurements result usually from a plethora of possible pathways that the solute can take and are, as such, an averaged out behavior, while in MD you mostly simulate just one pathway (the perpendicular crossing) without any added complexity (lipid patches that can be disordered, kinks and defects in the membranes, concentration gradients, etc).
Sorry the values are larger by 150 times
It still doesn’t change the nature of the question!