BACKGROUND: 8-Oxoguanine (8-oxoG) is a common oxidative lesion frequently encountered by DNA polymerases such as the repair enzyme DNA polymerase beta (pol beta). To interpret in atomic and energetic detail how pol beta processes 8-oxoG, we apply transition path sampling to delineate closing pathways of pol beta 8-oxoG complexes with dCTP and dATP incoming nucleotides and compare the results to those of the nonlesioned G:dCTP and G:dATPanalogues. RESULTS: Our analyses show that the closing pathways of the 8-oxoG complexes are different from one another and from the nonlesioned analogues in terms of the individual transition states along each pathway, associated energies, and the stability of each pathway's closed state relative to the corresponding open state. In particular, the closed-to-open state stability difference in each system establishes a hierarchy of stability (from high to low) as G:C > 8-oxoG:C > 8-oxoG:A > G:A, corresponding to -3, -2, 2, 9 kBT, respectively. This hierarchy of closed state stability parallels the experimentally observed processing efficiencies for the four pairs. Network models based on the calculated rate constants in each pathway indicate that the closed species are more populated than the open species for 8-oxoG:dCTP, whereas the opposite is true for 8-oxoG:dATP. CONCLUSION: These results suggest that the lower insertion efficiency (larger Km) for dATP compared to dCTP opposite 8-oxoG is caused by a less stable closed-form of pol beta, destabilized by unfavorable interactions between Tyr271 and the mispair. This stability of the closed vs. open form can also explain the higher insertion efficiency for 8-oxoG:dATP compared to the nonlesioned G:dATP pair, which also has a higher overall conformational barrier. Our study offers atomic details of the complexes at different states, in addition to helping interpret the different insertion efficiencies of dATP and dCTP opposite 8-oxoG and G.