CHARMM is an academic research program used widely for macromolecular mechanics and dynamics with versatile analysis manipulation tools of atomic coordinates trajectories. CHARMM-GUI, http://www.charmm-gui.org, has been developed to provide a web-based graphical user interface generate various input files molecular systems facilitate standardize the usage common advanced simulation techniques in CHARMM. The web environment provides ideal platform build validate model system interactive...

Molecular dynamics simulations of membrane proteins have provided deeper insights into their functions and interactions with surrounding environments at the atomic level. However, compared to solvation globular proteins, building a realistic protein/membrane complex is still challenging requires considerable experience simulation software. Membrane Builder in CHARMM-GUI website (http://www.charmm-gui.org) helps users build such system using web browser graphical user interface. Through...

CHARMM36 (C36) is the most up-to-date pairwise additive all-atom lipid force field and able to accurately represent bilayer properties of saturated monounsaturated molecules in natural constant particle, pressure, temperature (NPT) ensemble. However, molecular dynamics (MD) simulations on 1-stearoyl-2-docosahexaenoyl-sn-glycerco-3-phosphocholine (SDPC) bilayers polyunsaturated fatty acid (PUFA) chains result inaccuracies surface area per (SA), deuterium order parameters (S(CD)), X-ray form...

In this paper, we investigate the problem of overfitting in deep reinforcement learning. Among most common benchmarks RL, it is customary to use same environments for both training and testing. This practice offers relatively little insight into an agent's ability generalize. We address issue by using procedurally generated construct distinct test sets. Most notably, introduce a new environment called CoinRun, designed as benchmark generalization RL. Using find that agents overfit...

Traditional CPUs and cloud systems based on them have embraced the hardware-based trusted execution environments to securely isolate computation from malicious OS or hardware attacks. However, GPUs their deployments yet include such support for computing. As large amounts of sensitive data are offloaded GPU acceleration in environments, ensuring security is a current pressing need. deployed today, outsourced model vulnerable attacks compromised privileged software. To isolated remote even...

The shielded computation of hardware-based trusted execution environments such as Intel Software Guard Extensions (SGX) can provide secure cloud computing on remote systems under untrusted privileged system software. However, hardware overheads for securing protected memory restrict its capacity to a modest size several tens megabytes, and more demands beyond the limit cause costly demand paging. Although one widely used applications benefiting from enhanced security SGX, is in-memory...

A method of window exchange umbrella sampling molecular dynamics simulation is employed for transmembrane helix assembly. An analytical expression the average acceptance probability between neighboring windows derived and combined with first passage time optimization to predetermine a parameter set in an optimal range. With set, provides substantially more efficient helix-helix interfaces together potential mean force along distance helix-dimer model, compared method.

// Jinlong Yin 1,* , Gunwoo Park Jeong Eun Lee 2,3 Ju Young Tae-Hoon Kim 1 Youn-Jae Seung-Hoon 1,3 Heon Yoo Jong and Bae Specific Organs Cancer Branch, Research Institute, National Center, Goyang, Gyeonggi Korea 2 Cell Molecular Biology 3 Department of System Science, Graduate School Science Policy, * These authors contributed equally to this work Correspondence: Park, email: Kim, Yoo, Keywords : Glioma stem cell, self-renewal, differentiation, CPEB1 Received June 26, 2014 Accepted July 22,...

Although hardware-based trusted execution environments (TEEs) have evolved to provide strong isolation with efficient hardware supports, their current monolithic model poses challenges in representing common software structures modules produced from potentially untrusted 3rd parties. For better mapping of such modular designs environments, it is necessary extend the a hierarchical one, which provides multiple inner TEEs within TEE. compartmentalization TEE, this paper proposes novel TEE...

To prevent physical attacks on systems, secure processors have been proposed to reduce trusted computing base the processor itself. In a processor, all off-chip data are encrypted and their integrity is protected. This paper investigates how limited memory bandwidth of multi-core affects design processors. Although performance single-core has improved significantly with counter-mode encryption combined Bonsai Merkle Tree, our results indicate that can suffer from significant degradation due...

Finding relevant articles from PubMed is challenging because it hard to express the user's specific intention in given query interface, and a keyword typically retrieves large number of results. Researchers have applied machine learning techniques find by ranking according learned relevance function. However, process usually done offline without integrated with queries, users provide amount training documents get reasonable accuracy. This paper proposes novel multi-level feedback system for...

We propose to learn a curriculum or syllabus for supervised learning and deep reinforcement with neural networks by an attachable network, called ScreenerNet. Specifically, we weight each sample jointly training the ScreenerNet main network in end-to-end self-paced fashion. The neither has sampling bias nor requires remember past history. show augmented achieve early convergence better accuracy than state-of-the-art curricular methods extensive experiments using three popular vision datasets...

Delayed dark adaptation due to impaired rod photoreceptor homeostasis has been reported as the earliest symptom of eye diseases such age-related macular degeneration, diabetic retinopathy, and retinitis pigmentosa. Objective measurement can facilitate early diagnosis enable prompt intervention prevent vision loss. However, there is a lack noninvasive methods capable spatiotemporal monitoring changes during adaptation. Here we demonstrate functional optical coherence tomography (OCT) for in...

Phospholamban (PLB) is an integral membrane protein of 52 residues that regulates the activity sarcoplasmic reticulum calcium pump in cardiac muscle cells through reversible phosphorylation Ser16. To explore its possible conformations and dynamics a monomeric state, we have performed comparative molecular simulations unphosphorylated phosphorylated PLB (pPLB) with various orientations POPC membranes. The indicate cytoplasmic domain highly dependent on interactions membranes, is, large...

Abstract Understanding the delicate balance of forces governing helix or β‐hairpin interactions in transmembrane (TM) proteins is central to understanding membrane structure and function. These constituent play an essential role determining function proteins, protein membranes, thus form basis for many vital processes, including TM signaling, transport ions small molecules, energy transduction, cell–cell recognition. “Why does a single‐pass have specific orientations membranes?” “What are...

The increasing memory requirements of big data applications have been driving the precipitous growth capacity in server systems. To maximize efficiency external memory, HW-based compression techniques proposed to increase effective capacity. Although such can improve significantly, a critical trade-off exists techniques. As blocks need be decompressed as quickly possible serve cache misses, latency-optimized apply at cacheline granularity, achieving decompression latency less than few...

In designing feedforward neural networks, one often trains a large network and then prunes less useful hidden units. this paper, two non-heuristic pruning algorithms are derived from the Schmidt procedure. both, orthonormal systems of basis functions found, ordered, pruned, mapped back to original network. first algorithm, found ordered at time. optimal pruning, best subset is for each size Linear dependency considered computational cost analyzed. Simulation results given.

Finding related articles from the PubMed (a large biomedical literature repository) is challenging because it hard to express user's specific relevance in given query interface and a keyword typically retrieves many results. Biomedical researchers spend critical amount of time (e.g., often more than several days) search process. This paper proposes RefMed, novel system for PubMed, which supports ranking by enabling feedback on PubMed. RefMed first returns initial result documents as The user...