Formal analysis of human-computer interaction using model checking

Experiments with simulators allow psychologists to better understand the causes of human errors and build models of cognitive processes to be used in human reliability assessment (HRA). This paper investigates an approach to task failure analysis based on patterns of behaviour, by contrast to more traditional event-based approaches. It considers, as a case study, a formal model of an air traffic control (ATC) system which incorporates controller behaviour. The cognitive model is formalised in the CSP process algebra. Patterns of behaviour are expressed as temporal logic properties. Then a model-checking technique is used to verify whether the decomposition of the operator's behaviour into patterns is sound and complete with respect to the cognitive model. The decomposition is shown to be incomplete and a new behavioural pattern is identified, which appears to have been overlooked in the analysis of the data provided by the experiments with the simulator. This illustrates how formal analysis of operator models can yield fresh insights into how failures may arise in interactive systems.

A participatory design agenda for ubiquitous computing and multimodal interaction: A case study of dental practice

This paper reflects upon our attempts to bring a participatory design approach to design research into interfaces that better support dental practice. The project brought together design researchers, general and specialist dental practitioners, the CEO of a dental software company and, to a limited extent, dental patients. We explored the potential for deployment of speech and gesture technologies in the challenging and authentic context of dental practices. The paper describes the various motivations behind the project, the negotiation of access and the development of the participant relationships as seen from the researchers' perspectives. Conducting participatory design sessions with busy professionals demands preparation, improvisation, and clarity of purpose. The paper describes how we identified what went well and when to shift tactics. The contribution of the paper is in its description of what we learned in bringing participatory design principles to a project that spanned technical research interests, commercial objectives and placing demands upon the time of skilled professionals. Copyright © 2010 ACM, Inc

Cross-reactive recognition of human and primate cytomegalovirus sequences by human CD4 cytotoxic T lymphocytes specific for glycoprotein B and H

Although the importance of CD4(+) T cell responses to human cytonnegalovirus (HCMV) has recently been recognized in transplant and immunosuppressed patients, the precise specificity and nature of this response has remained largely unresolved. In the present study we have isolated CD4(+) CTL which recognize epitopes from HCMV glycoproteins gB and gH in association with two different HLA-DR antigens, DRA1*0101/DRB1*0701 (DR7) and DRA1*0101/DRB1*1101 (DR11). Comparison of amino acid sequences of HICMV isolates revealed that the gB and gH epitope sequences recognized by human CD4(+) T cells were not only conserved in clinical isolates from HCMV but also in CMV isolates from higher primates (chimpanzee, rhesus and baboon). Interestingly, these epitope sequences from chimpanzee, rhesus and baboon CMV are efficiently recognized by human CD4(+) CTL. More importantly, we show that gB-specific T cells from humans can also efficiently lyse pepticle-sensitized Patr-DR7(+) cells from chimpanzees. These findings suggest that conserved gB and gH epitopes should be considered while designing a prophylactic vaccine against HCMV. In addition, they also provide a functional basis for the conservation of MHC class 11 lineages between humans and Old World primates and open the possibility for the use of such primate models in vaccine development against HCMV.

Capsule and fimbria interaction in Klebsiella pneumoniae

The capsular polysaccharide and type I fimbriae are two of the major surface-located virulence properties associated with the pathogenesis of Klebsiella pneumoniae. The capsule is an elaborate polysaccharide matrix that encases the entire cell surface and provides resistance against many host defense mechanisms. In contrast, type 1 fimbriae are thin adhesive thread-like surface organelles that can extend beyond the capsular matrix and mediate D-mannose-sensitive adhesion to host epithelial cells. These fimbriae are archetypical and consist of a major building block protein (FimA) that comprises the bulk of the organelle and a tip-located adhesin (FimH). It is assumed that the extended major-subunit protein structure permits the FimH adhesin to function independently of the presence of a capsule. In this study, we have employed a defined set of K. pneumoniae capsulated and noncapsulated strains to show that the function of type I fimbriae is actually impeded by the concomitant expression of a polysaccharide capsule. Capsule expression had significant effects on two parameters commonly used to define FimH function, namely, yeast cell agglutination and biofilm formation. Our data suggest that this effect is not due to transcriptional/translational changes in fimbrial gene/protein expression but rather the result of direct physical interference. This was further demonstrated by the fact that we could restore fimbrial function by inhibiting capsule synthesis. It remains to be determined whether the expression of these very different surface components occurs simply via random events of phase variation or in a coordinated manner in response to specific environmental cues.

Lead compounds for antimalarial chemotherapy: Purine base analogs discriminate between human and P-falciparum 6-oxopurine phosphoribosyltransferases

The malarial parasite Plasmodium falciparum depends on the purine salvage enzyme hypoxanthine-guanine-xanthine phosphoribosyltransferase (HGXPRT) to convert purine bases from the host to nucleotides needed for DNA and RNA synthesis. An approach to developing antimalarial drugs is to use HGXPRT to convert introduced purine base analogs to nucleotides that are toxic to the parasite. This strategy requires that these compounds be good substrates for the parasite enzyme but poor substrates for the human counterpart, HGPRT. Bases with a chlorine atom in the 6-position or a nitrogen in the 8-position exhibited strong discrimination between P. falciparum HGXPRT and human HGPRT. The k(cat)/K-m values for the Plasmodium enzyme using 6-chloroguanine and 8-azaguanine as substrates were 50-80-fold and 336-fold higher than for the human enzyme, respectively. These and other bases were effective in inhibiting the growth of the parasite in vitro, giving IC50 values as low as 1 mu M.

Direct electrochemistry of human and rat NADPH cytochrome P450 reductase

The diflavo-protein NADPH cytochrome P450 reductase (CPR) is the key electron transfer partner for all drug metabolizing cytochrome P450 enzymes in humans. The protein delivers, consecutively, two electrons to the heme active site of the P450 in a carefully orchestrated process which ultimately leads to the generation of a high valent oxo-heme moiety. Despite its central role in P450 function, no direct electrochemical investigation of the purified protein has been reported. Here we report the first voltammetric study of purified human CPR where responses from both the FMN and FAD cofactors have been identified using both cyclic and square wave voltammetry. For human CPR redox responses at -2 and -278 mV (with a ratio of 1e(-):3e(-)) vs NHE were seen at pH 7.9 while the potentials for rat CPR at pH 8.0 were -20 and -254 mV. All redox responses exhibit a pH dependence of approximately -59 mV/pH unit consistent with proton coupled electron transfer reactions of equal stoichiometry. (c) 2006 Elsevier B.V. All rights reserved.