We combine single molecule force spectroscopy measurements with all-atom metadynamics simulations to investigate the cross-materials binding strength trends of DNA fragments adsorbed at the aqueous graphite C(0001) and Au(111) interfaces. Our simulations predict this adsorption at the level of the nucleobase, nucleoside, and nucleotide. We find that despite challenges in making clear, careful connections between the experimental and simulation data, reasonable consistency between the binding trends between the two approaches and two substrates was evident. On C(0001), our simulations predict a binding trend of dG > dA ≈ dT > dC, which broadly aligns with the experimental trend. On Au(111), the simulation-based binding strength trends reveal stronger adsorption for the purines relative to the pyrimadines, with dG ≈ dA > dT ≈ dC. Moreover, our simulations provide structural insights into the origins of the similarities and differences in adsorption of the nucleic acid fragments at the two interfaces. In particular, our simulation data offer an explanation for the differences observed in the relative binding trend between adenosine and guanine on the two substrates.

中文翻译：

---

通过实验与模拟相结合的方法在石墨和Au（111）上吸附DNA片段

我们将单分子力谱测量与全原子元动力学模拟相结合，以研究吸附在水性石墨C（0001）和Au（111）界面上的DNA片段的跨材料结合强度趋势。我们的模拟预测了在核碱基，核苷和核苷酸水平上的这种吸附。我们发现，尽管在实验数据和模拟数据之间建立清晰，仔细的联系时遇到了挑战，但两种方法与两种底物之间的结合趋势之间的合理一致性是显而易见的。在C（0001）上，我们的模拟预测dG> dA≈dT> dC的结合趋势，与实验趋势基本一致。在Au（111）上，基于模拟的结合强度趋势显示相对于嘧啶类化合物，嘌呤具有更强的吸附力，dG≈dA> dT≈dC。而且，我们的模拟为了解两个界面上核酸片段的吸附相似性和差异性的起源提供了结构上的见解。特别是，我们的模拟数据解释了在两种底物上腺苷和鸟嘌呤之间相对结合趋势中观察到的差异。