The physiological signature of sadness: A comparison between text, film and virtual reality.

Studies focused on the ubiquitous emotion of sadness demonstrate substantial variability in physiological responses during sadness elicitation, with no consensus regarding the physiological pattern of sadness. Variability in findings could be attributed to (a) the use of different induction techniques across studies or (b) the existence of subtypes of sadness with distinct physiological activation patterns. Typically, studies have used text and film to elicit sadness. However, virtual reality (VR) confers advantages over more traditional methods by allowing individuals a subjective sense of "being there" or presence. We compared participants' physiological responses to the same narrative presented via VR, Film and Story (n = 20 each) and collected their subjective responses to the stimuli. Results confirmed that participants in all conditions experienced the discrete emotion of sadness. Moreover, participants in the VR condition experienced the highest degree of presence. Regarding psychophysiological responses, participants in the VR condition had the lowest degree of baseline-adjusted parasympathetic activation in comparison to participants in the Film condition. Furthermore, while participants in the VR group showed diminished baseline-adjusted respiration rate and parasympathetic activation with an increase in presence, the opposite pattern was true for participants in the other conditions. The data suggest that the VR condition may elicit an activating pattern of sadness; whereas Film and Story conditions may elicit a deactivating pattern of sadness. Our results have implications for research using the discrete model of emotion, highlighting that different emotion elicitation techniques may result in differing expressions of what is considered a unitary emotion.

Moving least-squares reconstruction of large models with GPUs.

Modern laser range scanning campaigns produce extremely large point clouds, and reconstructing a triangulated surface thus requires both out-of-core techniques and significant computational power. We present a GPU-accelerated implementation of the moving least-squares (MLS) surface reconstruction technique. We believe this to be the first GPU-accelerated, out-of-core implementation of surface reconstruction that is suitable for laser range-scanned data. While several previous out-of-core approaches use a sweep-plane approach, we subdivide the space into cubic regions that are processed independently. This independence allows the algorithm to be parallelized using multiple GPUs, either in a single machine or a cluster. It also allows data sets with billions of point samples to be processed on a standard desktop PC. We show that our implementation is an order of magnitude faster than a CPU-based implementation when using a single GPU, and scales well to 8 GPUs.

Efficient compression of molecular dynamics trajectory files.

We investigate whether specific properties of molecular dynamics trajectory files can be exploited to achieve effective file compression. We explore two classes of lossy, quantized compression scheme: "interframe" predictors, which exploit temporal coherence between successive frames in a simulation, and more complex "intraframe" schemes, which compress each frame independently. Our interframe predictors are fast, memory-efficient and well suited to on-the-fly compression of massive simulation data sets, and significantly outperform the benchmark BZip2 application. Our schemes are configurable: atomic positional accuracy can be sacrificed to achieve greater compression. For high fidelity compression, our linear interframe predictor gives the best results at very little computational cost: at moderate levels of approximation (12-bit quantization, maximum error ≈ 10(-2) Å), we can compress a 1-2 fs trajectory file to 5-8% of its original size. For 200 fs time steps-typically used in fine grained water diffusion experiments-we can compress files to ~25% of their input size, still substantially better than BZip2. While compression performance degrades with high levels of quantization, the simulation error is typically much greater than the associated approximation error in such cases.

Simulation of Coarse-Grained Protein-Protein Interactions with Graphics Processing Units.

We report a hybrid parallel central and graphics processing units (CPU-GPU) implementation of a coarse-grained model for replica exchange Monte Carlo (REMC) simulations of protein assemblies. We describe the design, optimization, validation, and benchmarking of our algorithms, particularly the parallelization strategy, which is specific to the requirements of GPU hardware. Performance evaluation of our hybrid implementation shows scaled speedup as compared to a single-core CPU; reference simulations of small 100 residue proteins have a modest speedup of 4, while large simulations with thousands of residues are up to 1400 times faster. Importantly, the combination of coarse-grained models with highly parallel GPU hardware vastly increases the length- and time-scales accessible for protein simulation, making it possible to simulate much larger systems of interacting proteins than have previously been attempted. As a first step toward the simulation of the assembly of an entire viral capsid, we have demonstrated that the chosen coarse-grained model, together with REMC sampling, is capable of identifying the correctly bound structure, for a pair of fragments from the human hepatitis B virus capsid. Our parallel solution can easily be generalized to other interaction functions and other types of macromolecules and has implications for the parallelization of similar N-body problems that require random access lookups.

Visualisation of cyclic and multi-branched molecules with VMD.

We report the addition of two visualisation algorithms, termed PaperChain and Twister, to the freely available Visual Molecular Dynamics (VMD) package. These algorithms produce visualisations of complex cyclic molecules and multi-branched polysaccharides and are a generalization and optimization of those we previously developed in a standalone package for carbohydrates. PaperChain highlights each ring in a molecular structure with a polygon, which is coloured according to the ring pucker. Twister traces glycosidic bonds with a ribbon that twists according to the relative orientation of successive sugar residues. Combination of these novel algorithms and new ring selection statements with the large set of visualisations already available in VMD allows for unprecedented flexibility in the level of detail displayed for glycoconjugate, glycoprotein and carbohydrate-binding protein structures, as well as other cyclic structures. We highlight the efficacy of these algorithms with selected illustrative examples, clearly demonstrating the value of the new visualisations, not only for structure validation, but also for facilitating insights into molecular structure and mechanism.

Comparing the accuracy and precision of three techniques used for estimating missing landmarks when reconstructing fossil hominin crania.

Various methodological approaches have been used for reconstructing fossil hominin remains in order to increase sample sizes and to better understand morphological variation. Among these, morphometric quantitative techniques for reconstruction are increasingly common. Here we compare the accuracy of three approaches--mean substitution, thin plate splines, and multiple linear regression--for estimating missing landmarks of damaged fossil specimens. Comparisons are made varying the number of missing landmarks, sample sizes, and the reference species of the population used to perform the estimation. The testing is performed on landmark data from individuals of Homo sapiens, Pan troglodytes and Gorilla gorilla, and nine hominin fossil specimens. Results suggest that when a small, same-species fossil reference sample is available to guide reconstructions, thin plate spline approaches perform best. However, if no such sample is available (or if the species of the damaged individual is uncertain), estimates of missing morphology based on a single individual (or even a small sample) of close taxonomic affinity are less accurate than those based on a large sample of individuals drawn from more distantly related extant populations using a technique (such as a regression method) able to leverage the information (e.g., variation/covariation patterning) contained in this large sample. Thin plate splines also show an unexpectedly large amount of error in estimating landmarks, especially over large areas. Recommendations are made for estimating missing landmarks under various scenarios.

Techniques for visualization of carbohydrate molecules.

Standard molecular visualizations, such as the classic ball-and-stick model, are not suitable for large, complex molecules because the overall molecular structure is obscured by the atomic detail. For proteins, the more abstract ribbon and cartoon representations are instead used to reveal large scale molecular conformation and connectivity. However, there is currently no accepted convention for simplifying oligo- and polysaccharide structures. We introduce two novel visualization algorithms for carbohydrates, incorporated into a visualization package, CarboHydra. Both algorithms highlight the sugar rings and backbone conformation of the carbohydrate chain, ignoring ring substituents. The first algorithm, termed PaperChain, emphasizes the type and conformation of the carbohydrate rings. The second, Twister, emphasizes the relative orientation of the rings. We further include two rendering enhancements to augment these visualizations: silhouettes edges and a translucent overlay of the ball-and-stick atomic representation. To demonstrate their utility, the algorithms and visualization enhancements are here applied to a variety of carbohydrate molecules. User evaluations indicate that they present a more useful view of carbohydrate structure than the standard ball-and-stick representation. The algorithms were found to be complementary, with PaperChain particularly effective for smaller carbohydrates and Twister useful at larger scales for highlighting the backbone twist of polysaccharides.

Warp sculpting.

The task of computer-based free-form shape design is fraught with practical and conceptual difficulties. Incorporating elements of traditional clay sculpting has long been recognized as a means of shielding the user from these complexities. We present warp sculpting, a variant of spatial deformation, which allows deformations to be initiated by the rigid body transformation or uniform scaling of volumetric tools. This is reminiscent of a tool imprinting, flexing, and molding clay. Unlike previous approaches, the deformation is truly interactive. Tools, encoded in a distance field, can have arbitrarily complex shapes. Although individual tools have a static shape, several tools can be applied simultaneously. We enhance the basic formulation of warp sculpting in two ways. First, deformation is toggled to automatically overcome the problem of "sticky" tools, where the object's surface clings to parts of a tool that are moving away. Second, unlike many other spatial deformations, we ensure that warp sculpting remains foldover-free and, hence, prevent self-intersecting objects.