TOWARDS TRUE SSTABILIZATION ENERGIES OF H-BONDED AND STACKED DNA BASE PAIRS
Petr Jureèka and Pavel Hobza
Research Center for Complex Molecular Systems and Biomolecules,
Jaroslav Heyrovský Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, 18223 Praha
The study of binding energies of Adenine...Thymine and Guanine...Cytosine base pairs in vacuo was aimed to get an information on relative order of different structure motives on the energetical scale with special focuse on differences between H-bonded and stacked structures. We also aimed to find a general way to reliable interaction energies of the weakly bonded complexes.
Structures taken from previous MD/quench studies were fully optimized at RI-MP2 [1,2] level with TZVPP [5s3p2d1f]/[3s2p1d] basis set. To approach complete basis set (CBS) limit, convergency of both HF energy and MP2 correlation energy was studied employing augmented correlation-consistent basis sets aug-cc-pV(D,T,Q)Z. Is was found that molecular interaction energies close to the CBS limit may be obtained by 2-point extrapolation  using aug-cc-pVDZ and aug-cc-pVTZ basis sets.
To account for higher order correlation effects convergency of CCSD(T)→MP2 correction term (ΔEcorrMP2 - ΔEcorrCCSD(T)) was investigated. For the formamide….formamidine complex (Fig. 1) which is a model for adenine..thymine interaction MP2 and CCSD(T) correlation interaction energies were evaluated with various basis sets up to aug-cc-pVTZ (Fig. 2). It was shown that unlike the correlation energy itself the CCSD(T)→MP2 difference is almost basis set independent. Rather accurate values were obtained with relatively small 6-31G*(0.25) and cc-pVDZ(0.25,0.15) basis sets. Because the latter one performs well also for stacked complexes  it can be recommended for evaluation of the Δ term of extended complexes possessing both H-bonded and stacked structures.
Interaction energies of the DNA base pairs obtained by combination of the CBS extrapolations of MP2 interaction energies and the CCSD(T)→MP2 correction are in good agreement with experiment.
Fig. 1 Fig.2
1) M. Fayereisen, G. Fitzgerald & A. Komornicki, Chem. Phys. Lett., 208 (1993), 359.
2) Ahlrichs R., Bär M.& Häser M., Chem. Phys. Lett., 162 (1989), 165.
3) A. Halkier, T. Helgaker & P. Jørgensen, Chem. Phys. Lett., 302 (1999), 437-446.
4) Hobza P. & Šponer J., Chem. Phys. Lett., 288 (1998), 7-14.