اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدA deep learning approach to the structural analysis of proteins
76
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
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.
قيم البحث
اقرأ أيضاً
We explore the dynamical large-deviations of a lattice heteropolymer model of a protein by means of path sampling of trajectories. We uncover the existence of non-equilibrium dynamical phase-transitions in ensembles of trajectories between active and
inactive dynamical phases, whose nature depends on properties of the interaction potential. When the full heterogeneity of interactions due to the amino-acid sequence is preserved, as in a fully interacting model or in a heterogeneous version of the G={o} model where only native interactions are considered, the transition is between the equilibrium native state and a highly native but kinetically trapped state. In contrast, for the homogeneous G={o} model, where there is a single native energy and the sequence plays no role, the dynamical transition is a direct consequence of the static bi-stability between unfolded and native states. In the heterogeneous case the native-active and native-inactive states, despite their static similarity, have widely varying dynamical properties, and the transition between them occurs even in lattice proteins whose sequences are designed to make them optimal folders.
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.
A theoretical analysis of the unfolding pathway of simple modular proteins in length- controlled pulling experiments is put forward. Within this framework, we predict the first module to unfold in a chain of identical units, emphasizing the ranges of
pulling speeds in which we expect our theory to hold. These theoretical predictions are checked by means of steered molecular dynamics of a simple construct, specifically a chain composed of two coiled-coils motives, where anisotropic features are revealed. These simulations also allow us to give an estimate for the range of pulling velocities in which our theoretical approach is valid.
The convergent interests of different scientific disciplines, from biochemistry to electronics, toward the investigation of protein electrical properties, has promoted the development of a novel bailiwick, the so called proteotronics. The main aim of
proteotronics is to propose and achieve innovative electronic devices, based on the selective action of specific proteins. This paper gives a sketch of the fields of applications of proteotronics, by using as significant example the detection of a specific odorant molecule carried out by an olfactory receptor. The experiment is briefly reviewed and its theoretical interpretation given. Further experiments are envisioned and expected results discussed in the perspective of an experimental validation.
Neutron and X-ray reflectometry are powerful techniques facilitating the study of the structure of interfacial materials. The analysis of these techniques is ill-posed in nature requiring the application of a model-dependent methods. This can lead to
the over- and under- analysis of experimental data, when too many or too few parameters are allowed to vary in the model. In this work, we outline a robust and generic framework for the determination of the set of free parameters that is capable of maximising the in-formation density of the model. This framework involves the determination of the Bayesian evidence for each permutation of free parameters; and is applied to a simple phospholipid monolayer. We believe this framework should become an important component in reflectometry data analysis, and hope others more regularly consider the relative evidence for their analytical models.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
