scholarly journals DIRECT: RNA contact predictions by integrating structural patterns

2019 ◽  
Vol 20 (1) ◽  
Author(s):  
Yiren Jian ◽  
Xiaonan Wang ◽  
Jaidi Qiu ◽  
Huiwen Wang ◽  
Zhichao Liu ◽  
...  

Abstract Background It is widely believed that tertiary nucleotide-nucleotide interactions are essential in determining RNA structure and function. Currently, direct coupling analysis (DCA) infers nucleotide contacts in a sequence from its homologous sequence alignment across different species. DCA and similar approaches that use sequence information alone typically yield a low accuracy, especially when the available homologous sequences are limited. Therefore, new methods for RNA structural contact inference are desirable because even a single correctly predicted tertiary contact can potentially make the difference between a correct and incorrectly predicted structure. Here we present a new method DIRECT (Direct Information REweighted by Contact Templates) that incorporates a Restricted Boltzmann Machine (RBM) to augment the information on sequence co-variations with structural features in contact inference. Results Benchmark tests demonstrate that DIRECT achieves better overall performance than DCA approaches. Compared to mfDCA and plmDCA, DIRECT produces a substantial increase of 41 and 18%, respectively, in accuracy on average for contact prediction. DIRECT improves predictions for long-range contacts and captures more tertiary structural features. Conclusions We developed a hybrid approach that incorporates a Restricted Boltzmann Machine (RBM) to augment the information on sequence co-variations with structural templates in contact inference. Our results demonstrate that DIRECT is able to improve the RNA contact prediction.

Entropy ◽  
2020 ◽  
Vol 22 (5) ◽  
pp. 538
Author(s):  
Sangchul Oh ◽  
Abdelkader Baggag ◽  
Hyunchul Nha

A restricted Boltzmann machine is a generative probabilistic graphic network. A probability of finding the network in a certain configuration is given by the Boltzmann distribution. Given training data, its learning is done by optimizing the parameters of the energy function of the network. In this paper, we analyze the training process of the restricted Boltzmann machine in the context of statistical physics. As an illustration, for small size bar-and-stripe patterns, we calculate thermodynamic quantities such as entropy, free energy, and internal energy as a function of the training epoch. We demonstrate the growth of the correlation between the visible and hidden layers via the subadditivity of entropies as the training proceeds. Using the Monte-Carlo simulation of trajectories of the visible and hidden vectors in the configuration space, we also calculate the distribution of the work done on the restricted Boltzmann machine by switching the parameters of the energy function. We discuss the Jarzynski equality which connects the path average of the exponential function of the work and the difference in free energies before and after training.


1983 ◽  
Vol 34 (3) ◽  
pp. 241 ◽  
Author(s):  
CW Ford

Stem cell walls of pangola grass (Digitaria decumbens) were ground to two particle sizes (c. 1 and 0.1 mm diameter), and incubated with cellulase (ex. Trichoderma viride) for varying times before and after delignification. Total cell walls finely ground (0.1 mm) with a Spex Shatterbox mill were initially degraded more rapidly (to 24 h) than delignified 1 mm particles. Thereafter the delignified material was solubilized to a greater extent. Subsequent specific determinations of cell wall polysaccharides indicated that delignification increased the rate of hemicellulose degradation to a greater extent than did particle size reduction, whereas the opposite was found for cellulose. The difference between delignified and Spex-ground residues, in terms of the amount of polysaccharide digested, was much greater for cellulose than hemicellulose. It is concluded that structural features play a more important role in limiting cellulase degradation of cellulose than does association with lignin, the reverse being so for hemicellulose.


2017 ◽  
Vol E100.C (12) ◽  
pp. 1118-1121 ◽  
Author(s):  
Yasushi FUKUDA ◽  
Zule XU ◽  
Takayuki KAWAHARA

2008 ◽  
Vol 1069 ◽  
Author(s):  
Ryoji Kosugi ◽  
Toyokazu Sakata ◽  
Yuuki Sakuma ◽  
Tsutomu Yatsuo ◽  
Hirofumi Matsuhata ◽  
...  

ABSTRACTIn practical use of the SiC power MOSFETs, further reduction of the channel resistance, high stability under harsh environments, and also, high product yield of large area devices are indispensable. Pn diodes with large chip area have been already reported with high fabrication yield, however, there is few reports in terms of the power MOSFETs. To clarify the difference between the simple pn diodes and power MOSFETs, we have fabricated four pn-type junction TEGs having the different structural features. Those pn junctions are close to the similar structure of DIMOS (Double-implanted MOS) step-by-step from the simple pn diodes. We have surveyed the V-I characteristics dependence on each structural features over the 2inch wafer. Before their fabrication, we formed grid patterns with numbering over the 2inch wafer, then performed the synchrotron x-ray topography observation. This enables the direct comparison the electrical and spectrographic characteristics of each pn junctions with the fingerprints of defects.Four structural features from TypeA to TypeD are as follows. TypeA is the most simple structure as same as the standard pn diodes formed by Al+ ion implantation (I/I), except that the Al+ I/I condition conforms to that of the p-well I/I in the DIMOS. The JTE structure was used for the edge termination on all junctions. While the TypeA consists of one p-type region, TypeB and TypeC consists of a lot of p-wells. The difference of Type B and C is a difference of the oxide between the adjacent p-wells. The oxide of TypeB consists of the thick field oxide, while that of TypeC consists of the thermal oxide corresponding to the gate oxide in the DIMOS. In the TypeD structure, n+ region corresponding to the source in the DIMOS was added by the P+ I/I. The TypeD is the same structure of the DIMOS, except that the gate and source contacts are shorted. The V-I measurements of the pn junctions are performed using the KEITHLEY 237 voltage source meters with semi-auto probe machine. An active area of the fabricated pn junctions TEGs are 150um2 and 1mm2. Concentration and thickness of the drift layer are 1e16cm−3 and 10um, respectively.In order to compare the V-I characteristics of fabricated pn junctions with their defects information that obtained from x-ray topography measurements directly, the grid patterns are formed before the fabrication. The grid patterns were formed over the 2inch wafer by the SiC etching. The synchrotron x-ray topography measurements are carried out at the Beam-Line 15C in Photon-Factory in High-Energy-Accelerator-Research-Organization. Three diffraction conditions, g=11-28, -1-128, and 1-108, are chosen in grazing-incidence geometry (improved Berg-Barrett method).In the presentation, the V-I characteristics mapping on the 2inch wafer for each pn junctions, and the comparison of V-I characteristics with x-ray topography will be reported.


Sign in / Sign up

Export Citation Format

Share Document