Week 7

So why were the structures not keeping their correct structures? In our energy function we assume bond lengths and bond angles to be constant. The way a molecule changes conformation is through torsion angle movements, translations and rotations. However, minimization is done on the cartesian coordinates of each atom. Our new energy function only contains non bonded interaction terms. When minimizing the atoms move however they want to find the local minimum of the non bonded interactions.

Now, how do we fix this? Two ideas: put in "fake" bond and bond angle terms so that the minimization does not mess with them too much or to try minimizing over dihedral coordinates. I implemented the "bond and bond angle way and David implemented the dihedral coordinate way. Both of them seemed to work ok, but I think we're going to go with the dihedral coordinates because it's assures no bond and bond angle movement.

David and Mark have been working on the searching of the conformation space of the ligand near the active site of the protein in order to find the docked conformation. Initial results look ok, but not great. i worked with David some this week to tune the searching procedure.

Week: 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10