Review papers by their claims

Key empirical / computational claims and evidence

• Primary experimental observation: RNA:DNA hybrids (G:C-rich) give ~10× higher single-molecule conductance than identical-sequence B-form dsDNA; length series show ln(G) decreases with length and yield βRNA≈0.31 Å−1 and βDNA≈0.20 Å−1 (interpreted as hopping-dominated) — direct STM-BJ data and histograms reported Li et al. JPCL 2016 experimental data
• Structural evidence: CD spectra confirm RNA:DNA adopts A-form signatures (negative peak near 210 nm) while dsDNA in buffer is B-form; forcing dsDNA into A-form (75% ethanol) raises its conductance to similar values as RNA:DNA, showing structure (not base chemistry) dominates the effect Li et al. JPCL 2016 CD & control
• Computational support: DFT (B3LYP/6-31G(d,p)) + Green's-function DoS calculations show HOMO isosurface more spatially delocalized in RNA:DNA (≈70% of length) vs dsDNA (≈50%); 2D DoS maps show a much larger DoS near HOMO in the central bridge for RNA:DNA (ratios up to 10^4), which explains higher G but sparser accessible levels (fewer hopping sites) and thus larger β Li et al. JPCL 2016 DFT/DoS analysis

Critical appraisal — strengths, limitations, and blind spots

1. Strengths
   • High-throughput single-molecule STM-BJ statistics (thousands of traces) give robust conductance peaks and length-dependence trends (good experimental reproducibility) Li et al. JPCL 2016 STM statistics
   • Orthogonal CD + ethanol-induced A-form dsDNA control elegantly isolates conformation from base chemistry — a decisive experimental test supporting the structural hypothesis Li et al. JPCL 2016 conformation control
2. Limitations & uncertainties
   • Environmental/context dependence: STM-BJ measurements are performed with amine linkers to Au electrodes in buffered solution — junction geometry, electrode coupling (Γ), and linker chemistry strongly influence conductance and can bias β extraction; the authors used Γ=100 meV in DoS maps but real junction Γ can vary and changes interpretation of DoS-to-conductance mapping Li et al. JPCL 2016 Γ assumption
   • Small sequence set: study focuses on G:C-rich GGG–C(GC)n–GGG series — results may not generalize to A/T-rich sequences or mixed-sequence contexts; other reports show A-form vs B-form effects can be sequence-dependent Berlin et al. / Scientific Reports 2017 hole-transport
   • Modeling approximations: DFT level (B3LYP/6-31G(d,p)) with Kohn–Sham orbital projections captures qualitative delocalization but standard DFT often misplaces absolute orbital energies and level spacings; coupling to dynamic solvent and thermal fluctuations (conformational ensembles) is not fully sampled — the DoS maps are snapshots consistent with hypothesis but need ensemble averaging/MD + GW/OT-DFT corrections for quantitative energy spacing claims.
   • Interpretative leap: Authors argue 'fewer hopping sites' in RNA:DNA increases β — plausible, but hopping models require explicit calculation of rate constants (Marcus parameters), reorganization energies, and coupling between delocalized domains; the DoS sparsity argument is qualitative without explicit kinetic modeling of hops and inter-site coupling.
3. Potential biases to watch
   • Publication/positive-result bias: the paper finds a clear, internally consistent effect — but negative or contradictory sequences may be unpublished.
   • Linker/electrode bias: amine linkers chosen to maximize junction formation; different linkers (thiols) might change contact resistance and relative Gs.

Interpretation: how the data support the mechanistic picture

Concise chain-of-evidence (each step experimentally/computationally supported):

1. RNA:DNA hybrids conduct ≈10× more than identical dsDNA — robust STM histograms show single peaks across lengths (empirical) Li et al. JPCL 2016 conductance histograms
2. Conformation matters: CD and ethanol A-form control show A-form geometry correlates with high conductance (structure not chemistry) Li et al. JPCL 2016 CD control
3. Electronic structure explanation: DFT/DoS show HOMO delocalization across more bases + larger DoS near HOMO for RNA:DNA (higher conductance), but larger energy separation and fewer levels near HOMO (fewer hopping sites) produce larger β — consistent qualitative picture, though quantitative hopping rates are not computed explicitly Li et al. JPCL 2016 theory

Where could future work strengthen / falsify the interpretation?

• Direct hopping-rate computations: compute Marcus rates between delocalized domains using reorganization energies (λ), electronic coupling (V), and thermal activation (kBT) to test if sparser levels in RNA:DNA actually raise β while raising conductance at short lengths.
• MD ensemble averaging + electronic structure (GW or tuned-DFT) to quantify level spacings and thermal broadening in explicit solvent and ionic conditions; sample multiple junction-contact geometries to capture Γ variability.
• Sequence diversity: test A/T-rich, mixed sequences, and sequences with engineered disorder to see if A-form delocalization effect persists or is G-rich specific. Compare with photochemical hole-transport studies that show A-form hybrids can support delocalized HT across adenine tracts Hole transport in A-form hybrids (2017)
• Alternative contact chemistries: measure with thiol linkers, different electrode metals, and varying linker lengths to test robustness of 10× factor and β values.