A terabit capacity passive polymer optical backplane based on a novel meshed waveguide architecture

An optical backplane based on a meshed polymer waveguide architecture enabling high-speed board-to-board optical interconnection is presented. This planar array of multimode polymer waveguides can provide passive strictly non-blocking links between server line cards fitted with optical transmitter and receiver arrays. This architecture offers a scalable and low-cost solution to the bandwidth limitations faced by electrical backplanes and is suitable for PCB integration. The reported backplane demonstrator uses a matrix of 100 waveguides each capable of 10 Gb/s operation to interconnect 10 cards for a total capacity of a terabit per second aggregate data rate in multicast mode. Characterisation of the backplane demonstrator reveals low link losses of 2 to 8 dB for a multimode fibre input and crosstalk values below −35 dB. Error free data transmission at 10 Gb/s is achieved with a power penalty of only 0.2 dB at a bit-error-rate of 10⁻⁹. Additionally, lossless operation of a Gigabit Ethernet link over the backplane is achieved even when using the worst-case highest loss links.     

Application of machine learning to improve the results of high-throughput docking against the HIV-1 protease

We have previously reported that the application of a Laplacian-modified naive Bayesian (NB) classifier may be used to improve the ranking of known inhibitors from a random database of compounds after High-Throughput Docking (HTD). The method relies upon the frequency of substructural features among the active and inactive compounds from 2D fingerprint information of the compounds. Here we present an investigation of the role of extended connectivity fingerprints in training the NB classifier against HTD studies on the HIV-1 protease using three docking programs: Glide, FlexX, and GOLD. The results show that the performance of the NB classifier is due to the presence of a large number of features common to the set of known active compounds rather than a single structural or substructural scaffold. We demonstrate that the Laplacian-modified naive Bayesian classifier trained with data from high-throughput docking is superior at identifying active compounds from a target database in comparison to conventional two-dimensional substructure search methods alone.     

Sensitive, indirect photometric detector for high-performance liquid chromatography using a light-emitting diode

A low-noise detector for indirect photometric detection has been constructed using a highly stable source--a light-emitting diode (LED). Use of the detector is demonstrated for reversed-phase liquid chromatography by adding methylene blue to the mobile phase to make a background signal. The indirect determination of alcohols by their effect on methylene blue concentration distribution is demonstrated, and an investigation is made into the conditions for high sensitivity. Because the source exhibits low noise, the detection limits for alcohols are as low as more complex and expensive detection methods, despite the lower radiant power of the LED. Detection limits for nine alcohols are below micrograms injected amounts.     

Enrichment of high-throughput screening data with increasing levels of noise using support vector machines, recursive partitioning, and laplacian-modified naive bayesian classifiers

High-throughput screening (HTS) plays a pivotal role in lead discovery for the pharmaceutical industry. In tandem, cheminformatics approaches are employed to increase the probability of the identification of novel biologically active compounds by mining the HTS data. HTS data is notoriously noisy, and therefore, the selection of the optimal data mining method is important for the success of such an analysis. Here, we describe a retrospective analysis of four HTS data sets using three mining approaches: Laplacian-modified naive Bayes, recursive partitioning, and support vector machine (SVM) classifiers with increasing stochastic noise in the form of false positives and false negatives. All three of the data mining methods at hand tolerated increasing levels of false positives even when the ratio of misclassified compounds to true active compounds was 5:1 in the training set. False negatives in the ratio of 1:1 were tolerated as well. SVM outperformed the other two methods in capturing active compounds and scaffolds in the top 1%. A Murcko scaffold analysis could explain the differences in enrichments among the four data sets. This study demonstrates that data mining methods can add a true value to the screen even when the data is contaminated with a high level of stochastic noise.     

Design and characterization of polymers and polymer blends for high temperature structural applications

Reviewing the development of new polymeric materials for high temperature structural applications (T > 200°C) over the past several decades, reveals a paradox which, to date, has not been completely resolved. Polymers which exhibit very high temperature stability tend to be either intractable or brittle, whereas, easily processible polymers tend to fall short of property targets. Approaches to resolving this paradox include modification of the chain backbone chemistry and polymer blending (especially to form miscible systems). Recent research has shown that, in contrast to low temperature flexible polymers, many high temperature aromatic heterocyclic polymers form miscible systems which permit the design of the desired processibility and performance into the blend. An example of such a system is the blend of Poly(2,2′-(meta-phenylene-5,5′-bibenzimidazole) (PBI) with a series of polyamides, including commercially available polyether imide (PEI) and imide copolymers containing sulfone and fluorinated isopropylidene (6F) units. Other examples include all polyimide blends and blends of polyimides with polyethersulfone.