Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userPredicting the DNA Conductance using Deep Feed Forward Neural Network Model
368
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
Double-stranded DNA (dsDNA) has been established as an efficient medium for charge migration, bringing it to the forefront of the field of molecular electronics as well as biological research. The charge migration rate is controlled by the electronic couplings between the two nucleobases of DNA/RNA. These electronic couplings strongly depend on the intermolecular geometry and orientation. Estimating these electronic couplings for all the possible relative geometries of molecules using the computationally demanding first-principles calculations requires a lot of time as well as computation resources. In this article, we present a Machine Learning (ML) based model to calculate the electronic coupling between any two bases of dsDNA/dsRNA of any length and sequence and bypass the computationally expensive first-principles calculations. Using the Coulomb matrix representation which encodes the atomic identities and coordinates of the DNA base pairs to prepare the input dataset, we train a feedforward neural network model. Our NN model can predict the electronic couplings between dsDNA base pairs with any structural orientation with a MAE of less than 0.014 eV. We further use the NN predicted electronic coupling values to compute the dsDNA/dsRNA conductance.
rate research
Read More
In this work, we model the zero-bias conductance for the four different DNA strands that were used in conductance measurement experiment [A. K. Mahapatro, K. J. Jeong, G. U. Lee, and D. B. Janes, Nanotechnology 18, 195202 (2007)]. Our approach consists of three elements: (i) ab initio calculations of DNA, (ii) Greens function approach for transport calculations, and (iii) the use of two parameters to determine the decoherence rates. We first study the role of the backbone. We find that the backbone can alter the coherent transmission significantly at some energy points by interacting with the bases, though the overall shape of the transmission stays similar for the two cases. More importantly, we find that the coherent electrical conductance is tremendously smaller than what the experiments measure. We consider DNA strands under a variety of different experimental conditions and show that even in the most ideal cases, the calculated coherent conductance is much smaller than the experimental conductance. To understand the reasons for this, we carefully look at the effect of decoherence. By including decoherence, we show that our model can rationalize the measured conductance of the four strands, both qualitatively and quantitatively. We find that the effect of decoherence on G:C base pairs is crucial in getting agreement with the experiments. However, the decoherence on G:C base pairs alone does not explain the experimental conductance in strands containing a number of A:T base pairs. Including decoherence on A:T base pairs is also essential. By fitting the experimental trends and magnitudes in the conductance of the four different DNA molecules, we estimate for the first time that the deocherence rate is 6 meV for G:C and 1.5 meV for A:T base pairs.
Deep Learning (DL) algorithms hold great promise for applications in the field of computational biophysics. In fact, the vast amount of available molecular structures, as well as their notable complexity, constitutes an ideal context in which DL-based approaches can be profitably employed. To express the full potential of these techniques, though, it is a prerequisite to express the information contained in the molecules atomic positions and distances in a set of input quantities that the network can process. Many of the molecular descriptors devised insofar are effective and manageable for relatively small structures, but become complex and cumbersome for larger ones. Furthermore, most of them are defined locally, a feature that could represent a limit for those applications where global properties are of interest. Here, we build a deep learning architecture capable of predicting non-trivial and intrinsically global quantities, that is, the eigenvalues of a proteins lowest-energy fluctuation modes. This application represents a first, relatively simple test bed for the development of a neural network approach to the quantitative analysis of protein structures, and demonstrates unexpected use in the identification of mechanically relevant regions of the molecule.
We report a theoretical study of DNA flexibility and quantitatively predict the ring closure probability as a function of DNA contour length. Recent experimental studies show that the flexibility of short DNA fragments (as compared to the persistence length of DNA l_P~150 base pairs) cannot be described by the traditional worm-like chain (WLC) model, e.g., the observed ring closure probability is much higher than predicted. To explain these observations, DNA flexibility is investigated with explicit considerations of a new length scale l_D~10 base pairs, over which DNA local bend angles are correlated. In this correlated worm-like chain (C-WLC) model, a finite length correction term is analytically derived and the persistence length is found to be contour length dependent. While our model reduces to the traditional worm-like chain model when treating long DNA at length scales much larger than l_P, it predicts that DNA becomes much more flexible at shorter sizes, which helps explain recent cyclization measurements of short DNA fragments around 100 base pairs.
In living cells, proteins combine 3D bulk diffusion and 1D sliding along the DNA to reach a target faster. This process is known as facilitated diffusion, and we investigate its dynamics in the physiologically relevant case of confined DNA. The confining geometry and DNA elasticity are key parameters: we find that facilitated diffusion is most efficient inside an isotropic volume, and on a flexible polymer. By considering the typical copy numbers of proteins in vivo, we show that the speedup due to sliding becomes insensitive to fine tuning of parameters, rendering facilitated diffusion a robust mechanism to speed up intracellular diffusion-limited reactions. The parameter range we focus on is relevant for in vitro systems and for facilitated diffusion on yeast chromatin.
In this short note, a correction is made to the recently proposed solution [1] to a 1D biased diffusion model for linear DNA translocation and a new analysis will be given to the data in [1]. It was pointed out [2] by us recently that this 1D linear translocation model is equivalent to the one that was considered by Schrodinger [3] for the Enrenhaft-Millikan measurements [4,5] on electron charge. Here we apply Schrodingers first-passage-time distribution formula to the data set in [1]. It is found that Schrodingers formula can be used to describe the time distribution of DNA translocation in solid-state nanopores. These fittings yield two useful parameters: drift velocity of DNA translocation and diffusion constant of DNA inside the nanopore. The results suggest two regimes of DNA translocation: (I) at low voltages, there are clear deviations from Smoluchowskis linear law of electrophoresis [6] which we attribute to the entropic barrier effects; (II) at high voltages, the translocation velocity is a linear function of the applied electric field. In regime II, the apparent diffusion constant exhibits a quadratic dependence on applied electric field, suggesting a mechanism of Taylor dispersion effect likely due the electro-osmotic flow field in the nanopore channel. This analysis yields a dispersion-free diffusion constant value for the segment of DNA inside the nanopore which is in agreement with Stokes-Einstein theory quantitatively. The implication of Schrodingers formula for DNA sequencing is discussed.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
