Maria I. Andreou

Intein-mediated site-specific conjugation of Quantum Dots to proteins in vivo

We describe an intein based method to site-specifically conjugate Quantum Dots (QDs) to target proteins in vivo. This approach allows the covalent conjugation of any nanostructure and/or nanodevice to any protein and thus the targeting of such material to any intracellular compartment or signalling complex within the cells of the developing embryo. We genetically fused a pleckstrin-homology (PH) domain with the N-terminus half of a split intein (IN). The C-terminus half (IC) of the intein was conjugated to QDs in vitro. IC-QDs and RNA encoding PH-IN were microinjected into Xenopus embryos. In vivo intein-splicing resulted in fully functional QD-PH conjugates that could be monitored in real time within live embryos. Use of Near Infra Red (NIR)-emitting QDs allowed monitoring of QD-conjugates within the embryo at depths where EGFP is undetectable demonstrating the advantages of QDs for this type of experiment. In conclusion, we have developed a novel in vivo methodology for the site-specific conjugation of QDs and other artificial structures to target proteins in different intracellular compartments and signaling complexes.

Crowdsourced Trace Similarity with Smartphones

Smartphones are nowadays equipped with a number of sensors, such as WiFi, GPS, accelerometers, etc. This capability allows smartphone users to easily engage in crowdsourced computing services, which contribute to the solution of complex problems in a distributed manner. In this work, we leverage such a computing paradigm to solve efficiently the following problem: comparing a query trace Q against a crowd of traces generated and stored on distributed smartphones. Our proposed framework, coined SmartTrace+, provides an effective solution without disclosing any part of the crowd traces to the query processor. SmartTrace+, relies on an in-situ data storage model and intelligent top-K query processing algorithms that exploit distributed trajectory similarity measures, resilient to spatial and temporal noise, in order to derive the most relevant answers to Q. We evaluate our algorithms on both synthetic and real workloads. We describe our prototype system developed on the Android OS. The solution is deployed over our own SmartLab testbed of 25 smartphones. Our study reveals that computations over SmartTrace+ result in substantial energy conservation; in addition, results can be computed faster than competitive approaches.

Adaptive probabilistic flooding for Information Hovering in VANETs

Information Hovering applies in many applications in Vehicular Ad Hoc Networks, where useful information needs to be made available to all vehicles within a confined geographical area for a specific time interval. A straightforward approach is to have all vehicles within the hovering area exchange messages with each other. However, this method does not guarantee that all vehicles within the hovering area will receive the message due to potential partitioning of the network in areas with low traffic density and/or low market penetration rate. To alleviate this problem in this work we propose a scheme which is based on the application of epidemic routing within the hovering area and probabilistic flooding outside the hovering area. Informed vehicles outside the area can serve as information bridges towards partitioned uninformed areas thus leading to high reachability. A unique feature of the proposed protocol is that it is adaptive in the sense that the rebroadcast probability outside the hovering area is adaptively regulated based on estimates of the vehicle density within the hovering area. We evaluate the performance of the proposed scheme using VISSIM. The reference model used in all simulation experiments represents a section of the road network in the cities of Bellevue and Redmond in Washington. The obtained simulation results indicate that the proposed protocol is successful in satisfying its design objectives and that it outperforms other candidate hovering protocols.

Time based personalization for the moving user

Unlike desktop users, mobile users are a new and more demanding breed. Technology provided for the first group is often found lacking for the later. Personalization is such an example. The moving user differs from the desktop user in that his handheld device is truly personal. It roams with the user and allows him access to info and services at any given time from anywhere. As the moving user is not limited to a fix place and to a given time period, factors such as time and current experience becomes increasingly important for him. His context is now a function of time and experience and the goal of personalization is to match the local services to this context. In this paper we exploit the importance of time and experience in personalization for the moving user and present a system that anticipates and compensates the time-dependant shifting of user interests.

Using time and activity in personalization for the mobile user

Mobile clients present a new and more demanding breed of users. Solutions provided for the desktop users are often found inadequate to support this new breed of users. Personalization is such a solution. The moving user differs from the desktop user in that his handheld device is truly personal. It roams with the user and allows him access to info and services at any given time from anywhere. As the moving user is not bound to a fixed place and to a given time period, factors such as time and current experience becomes increasingly important for him. His context and preferences are now a function of time and experience and the goal of personalization is to match the local services to this time-depended preferences. In this paper we exploit the importance of time and experience in personalization for the moving user and present a system that anticipates and compensates the time-dependant shifting of user interests. A prototype system is implemented and our initial evaluation results indicate performance improvements over traditional personalization schemes that range up to 173%.

Information Hovering in Vehicular Ad-Hoc Networks

Information Hovering is a relatively new concept of information dissemination over a mobile set of peers. It naturally applies in many applications in Vehicular Ad-Hoc Networks, where useful information needs to be made available to all vehicles within a confined geographical area for a specific time interval. A straightforward approach, is to have all vehicles within the hovering area exchange messages with each other. However, this method does not guarantee that all vehicles within the hovering area will receive the message due to potential parti- tioning of the network in areas with low traffic density and/or low market penetration rate. In this work, we address the problem by applying probabilistic flooding schemes outside the hovering area. Informed vehicles outside the hovering area can serve as information bridges towards partitioned uninformed areas thus leading to high reachability. We consider a number of rebroadcast probability functions and we evaluate their performance using the microscopic simulation tool VISSIM. Our reference model represents a section of the road network in the cities of Bellevue and Redmond in Washington. The obtained results indicate that probabilistic flooding with a Gaussian like probability function outperforms other approaches by achieving high reachability values and a relatively small number of exchanged messages.

In Vivo, Site-Specific, Covalent Conjugation of Quantum Dots to Proteins via Split-Intein Splicing

The ability to target proteins with nanostructures and/or nanodevices in vivo is important for understanding and controlling their biological function. Quantum dots (QDs) serve as an ideal model nanostructure due to their superior optical properties that permit visual confirmation of in vivo targeting and localization and due to their potential as a bio-imaging tool. Here, we describe the site-specific covalent conjugation of quantum dots to target proteins in vivo using an intein-based method. Experimental procedure includes the following steps: (1) fusion of Pleckstrin-homology (PH) domains with the N-terminus half of a split intein (I(N)); (2) conjugation of the C-terminal (I(C)) intein-derived peptide to streptavidin-coated QDs in vitro; and (3) in vivo expression of PH-I(N) following microinjection of PH-I(N) RNA and I(C)-QDs into Xenopus embryos. Intein-splicing results in covalent conjugation of QDs with the C-terminus of the PH domain without interfering with protein localization or function. Such produced QD-PH conjugates could be monitored in real time within live embryos.The use of near infrared-emitting QDs allows monitoring of QD conjugates within the embryo at depths where EGFP is undetectable demonstrating the advantages of QDs for this type of experiment. The reported approach therefore allows the covalent conjugation of QDs or other similar nanostructures to proteins in vivo and the targeting of such nanomaterial to any intracellular compartment or signaling -complex within the cells of the developing embryo.

Generating and radiocoloring families of perfect graphs

In this work we experimentally study the min order Radiocoloring problem (RCP) on Chordal, Split and Permutation graphs, which are three basic families of perfect graphs. This problem asks to find an assignment using the minimum number of colors to the vertices of a given graph G, so that each pair of vertices which are at distance at most two apart in G have different colors. RCP is an NP-Complete problem on chordal and split graphs [4]. For each of the three families, there are upper bounds or/and approximation algorithms known for minimum number of colors needed to radiocolor such a graph [4,10]. We design and implement radiocoloring heuristics for graphs of above families, which are based on the greedy heuristic. Also, for each one of the above families, we investigate whether there exists graph instances requiring a number of colors in order to be radiocolored, close to the best known upper bound for the family. Towards this goal, we present a number generators that produce graphs of the above families that require either (i) a large number of colors (compared to the best upper bound), in order to be radiocolored, called “extremal” graphs or (ii) a small number of colors, called “non-extremal”instances. The experimental evaluation showed that random generated graph instances are in the most of the cases “non-extremal” graphs. Also, that greedy like heuristics performs very well in the most of the cases, especially for “non-extremal” graphs.

HIV-1 transmission networks across Cyprus (2010-2012).

A molecular epidemiology study of HIV-1 infection was conducted in one hundred diagnosed and untreated HIV-1-infected patients in Cyprus between 2010 and 2012, representing 65.4% of all the reported HIV-1 infections in Cyprus in this three-year period, using a previously defined enrolment strategy. Eighty-two patients were newly diagnosed (genotypic drug resistance testing within six months from diagnosis), and eighteen patients were HIV-1 diagnosed for a longer period or the diagnosis date was unknown. Phylogenetic trees of the pol sequences obtained in this study with reference sequences indicated that subtypes B and A1 were the most common subtypes present and accounted for 41.0 and 19.0% respectively, followed by subtype C (7.0%), F1 (8.0%), CRF02_AG (4.0%), A2 (2.0%), other circulating recombinant forms (CRFs) (7.0%) and unknown recombinant forms (URFs) (12%). Most of the newly-diagnosed study subjects were Cypriots (63%), males (78%) with median age 39 (Interquartile Range, IQR 33–48) reporting having sex with other men (MSM) (51%). A high rate of clustered transmission of subtype B drug-sensitive strains to reverse transcriptase and protease inhibitors was observed among MSM, twenty-eight out of forty-one MSM study subjects (68.0%) infected were implicated in five transmission clusters, two of which are sub-subtype A1 and three of which are subtype B strains. The two largest MSM subtype B clusters included nine and eight Cypriot men, respectively, living in all major cities in Cyprus. There were only three newly diagnosed patients with transmitted drug resistant HIV-1 strains, one study subject from the United Kingdom infected with subtype B strain and one from Romania with sub-subtype A2 strain, both with PI drug resistance mutation M46L and one from Greece with sub-subtype A1 with non-nucleoside reverse transcriptase inhibitors (NNRTI) drug resistance mutation K103N.

40LoVe and Samba are involved in Xenopus neural development and functionally distinct from hnRNP AB.

Heterogeneous nuclear ribonucleoproteins (hnRNPs) comprise a large group of modular RNA-binding proteins classified according to their conserved domains. This modular nature, coupled with a large choice of alternative splice variants generates functional diversity. Here, we investigate the biological differences between 40LoVe, its splice variant Samba and its pseudoallele hnRNP AB in neural development. Loss of function experiments lead to defects in neural development with reduction of eye size, which stem primarily from increased apoptosis and reduced proliferation in neural tissues. Despite very high homology between 40LoVe/Samba and hnRNP AB, these proteins display major differences in localization, which appear to be in part responsible for functional differences. Specifically, we show that the 40Love/Samba carboxy-terminal domain (GRD) enables nucleocytoplasmic shuttling behavior. This domain is slightly different in hnRNP AB, leading to nuclear-restricted localization. Finally, we show that shuttling is required for 40LoVe/Samba function in neural development.