Molecular Fractionation with Conjugate Caps (MFCC)
Introduced in 2003, MFCC 99 focused on fragment-based calculations of proteins. The original MFCC method uses the MFCC Fragmentation Method and the Simple Replacement capping method (or at least we assume it does since the manuscript only mentions adding hydrogen atoms). The resulting fragments intersect with the fragments they are bonded to. Instead of directly using the IEP, MFCC corrects for intersections by combining the intersections into molecules (for each non-terminal amino acid, the left and right intersections are combined into a molecule). Initial small basis SCF and DFT (with the B3LYP functional) results focused on water-peptide interactions and were shown to be in excellent agreement with supersystem results.
In a follow up study Zhang et al.97 considered how the cap choice and cutting location affected the error by looking at the potential energy surface for the interaction of a water molecule with a five amino acid peptide. Calculations were at the SCF and B3LYP level of theory, with small basis sets (3-21G, 3-21+G, 3-21G*, and 6-31G). Results indicated that cutting the bond between the alpha-carbon and the carbonyl carbon, was better than cutting the peptide bond (depending on the particular potential energy surface point, could be more than 1 kcal/mol difference). Cap size was found to have a smaller effect on the error (about 0.5 kcal/mol).
The next development came when Chen et al.4 applied MFCC to proteins containing disulfide bonds. In addition to severing the alpha carbon-carbonyl carbon bonds, the sulfide bonds were also severed. They considered two capping strategies for the severed sulfide bonds: hydrogen atoms, or methyls. The latter was found to afford better results. Sample calculations were at the SCF, B3LYP, and MP2 levels of theory paired with several small basis sets (at most 6-31G).
Name |
Fragments |
Embedding |
Caps |
Screening |
---|---|---|---|---|
MFCC |
N/A |
N/A |
Other MFCC Results
Zhang et al.98 applied MFCC to a streptavidin-biotin system, which is a 121 amino-acid protein. Results are computed using the HF/3-21G model chemistry and compared to the AMBER force field. Relatively large energetic discrepancies between the two are noted.