Characterizing protein interactions employing a genome-wide siRNA cellular phenotyping screen.

Characterizing the activating and inhibiting effect of protein-protein interactions (PPI) is fundamental to gain insight into the complex signaling system of a human cell. A plethora of methods has been suggested to infer PPI from data on a large scale, but none of them is able to characterize the effect of this interaction. Here, we present a novel computational development that employs mitotic phenotypes of a genome-wide RNAi knockdown screen and enables identifying the activating and inhibiting effects of PPIs. Exemplarily, we applied our technique to a knockdown screen of HeLa cells cultivated at standard conditions. Using a machine learning approach, we obtained high accuracy (82% AUC of the receiver operating characteristics) by cross-validation using 6,870 known activating and inhibiting PPIs as gold standard. We predicted de novo unknown activating and inhibiting effects for 1,954 PPIs in HeLa cells covering the ten major signaling pathways of the Kyoto Encyclopedia of Genes and Genomes, and made these predictions publicly available in a database. We finally demonstrate that the predicted effects can be used to cluster knockdown genes of similar biological processes in coherent subgroups. The characterization of the activating or inhibiting effect of individual PPIs opens up new perspectives for the interpretation of large datasets of PPIs and thus considerably increases the value of PPIs as an integrated resource for studying the detailed function of signaling pathways of the cellular system of interest.

An aneurysm clipping training module for the neurosurgical training simulator NeuroSim.

Having introduced NeuroSim, the prototype of a neurosurgical training simulator at MMVR18, we present our first medical training module. NeuroSim is based on virtual reality and uses real-time algorithms for simulating tissue. It provides a native interface by using a real surgical microscope and original instruments. Having implemented some abstract tasks to train basic skills like hand-eye coordination or the handling of the microscope last year, we now present a medical module where an aneurysm has to be clipped. NeuroSim has been developed in cooperation with the neurosurgical clinic of the University of Heidelberg and VRmagic GmbH in Mannheim.

NeuroSim--the prototype of a neurosurgical training simulator.

We present NeuroSim, the prototype of a training simulator for open surgical interventions on the human brain. The simulator is based on virtual reality and uses real-time simulation algorithms to interact with models generated from MRT- or CT-datasets. NeuroSim provides a native interface by using a real surgical microscope and original instruments tracked by a combination of inertial measurement units and optical tracking. Conclusively an immersive environment is generated. In a first step the navigation in an open surgery setup as well as the hand-eye coordination through a microscope can be trained. Due to its modular design further training modules and extensions can be integrated. NeuroSim has been developed in cooperation with the neurosurgical clinic of the University of Heidelberg and the VRmagic GmbH in Mannheim.

EYESi ophthalmoscope - a simulator for indirect ophthalmoscopic examinations.

We present a training simulator for indirect ophthalmoscopy. An optical tracking system is used to reconstruct the position of a lens mockup and a model of the patient's face. Refraction and illumination are computed in real-time and displayed on a head-mounted display using augmented reality. A case database completes the training system which allows to practise the examination and to study clinical patterns.

Development of a guiding endoscopy simulator.

Endoscopy simulators get more and more common for the training of physicians. It is important to make simulation as realistic as possible by providing optical, acoustical and haptical feedback. The haptic display of our simulator EndoSim allows applying active forces to all degrees of freedom and moving to defined positions. This positioning is used for our automatic guiding system. If the user asks for help, an algorithm calculates how to get over the next barrier, factoring forces and distances. The system is able to decide if it is wise to choose a longer way in order to reduce the force. The user gets either an optical help shown by signs or is guided directly by the automatically moved endoscope. This guiding system is a new possibility for teaching physicians to increase their examination capabilities.

Real-time Marker-based Tracking of a Non-rigid Object.

Real-time tracking of non-rigid objects for use in interfaces of VR-simulators is presented. Markers are attached to the objects and observed by several cameras with integrated image-processing hardware which extracts relevant marker data (centroid, area & color) in real-time. Data from the different cameras is then matched in the host PC to reconstruct the 3D positions. We present two approaches to this special matching problem because standard image feature based algorithms are not feasible for marker-based tracking. A model of the deformation is extracted from the reconstructed 3D point cloud and the simulation model is updated accordingly. Experiments with a prototype of a deformable eye interface for the ophthalmosurgical simulator EYESI show that latency, robustness and accuracy of the deformation tracking are adequate for application in VR simulators. The approach is extensible to other types of simulators where deformable tissue has to be tracked.

Ultrasound, a new tool for surface matching in computer-navigated surgery.

OBJECTIVE: The object of this study was to investigate the feasibility of generating a bone surface from data provided by an ultrasound examination and to match this surface with the previous computed tomography (CT) scan. METHODS: From a CT data set of a training model of the pelvis, a three-dimensional surface was extracted by global thresholding-based segmentation. The same model was placed in a water basin, and ultrasound images were taken with a guided ultrasound transducer. The three-dimensional surface was generated from the ultrasound data set, and the two surfaces were matched in a semiautomatic mode. RESULTS: With special segmentation methods, a surface could be extracted automatically from the CT and the ultrasound data set. From these segmented ultrasound slices, a volume data set of the model was generated. After approximate initial matching, the local matching process was completed automatically. CONCLUSION: One of the limitations in computer-assisted surgery is the complicated matching process. Using special algorithms, a surface was extracted from the data set of an ultrasound examination and matched in a semiautomatic mode with the surface of a CT data set, facilitating the matching process.

A divide and conquer approach to fast loop modeling.

We describe a fast ab initio method for modeling local segments in protein structures. The algorithm is based on a divide and conquer approach and uses a database of precalculated look-up tables, which represent a large set of possible conformations for loop segments of variable length. The target loop is recursively decomposed until the resulting conformations are small enough to be compiled analytically. The algorithm, which is not restricted to any specific loop length, generates a ranked set of loop conformations in 20-180 s on a desktop PC. The prediction quality is evaluated in terms of global RMSD. Depending on loop length the top prediction varies between 1.06 A RMSD for three-residue loops and 3.72 A RMSD for eight-residue loops. Due to its speed the method may also be useful to generate alternative starting conformations for complex simulations.

Computer-generated stratified diffractive optical elements.

We present what is to our knowledge a new type of diffractive optical element (DOE), the computer-generated stratified diffractive optical element (SDOE), a hybridization of thin computer-generated DOEs and volume holograms. A model and several algorithms for calculating computer-generated SDOEs are given. Simulations and experimental results are presented that exhibit the properties of computer-generated SDOEs: the strong angular and wavelength selectivity of SDOEs makes it possible to store multiple pages in a computer-generated SDOE, which can be read out separately (multiplexing). The reconstruction of an optimized SDOE has a higher quality than the reconstruction of optimized one-layer DOEs. SDOEs can be calculated to have only one diffraction order.