Karen Duca

An insight-based methodology for evaluating bioinformatics visualizations

High-throughput experiments, such as gene expression microarrays in the life sciences, result in very large data sets. In response, a wide variety of visualization tools have been created to facilitate data analysis. A primary purpose of these tools is to provide biologically relevant insight into the data. Typically, visualizations are evaluated in controlled studies that measure user performance on predetermined tasks or using heuristics and expert reviews. To evaluate and rank bioinformatics visualizations based on real-world data analysis scenarios, we developed a more relevant evaluation method that focuses on data insight. This paper presents several characteristics of insight that enabled us to recognize and quantify it in open-ended user tests. Using these characteristics, we evaluated five microarray visualization tools on the amount and types of insight they provide and the time it takes to acquire it. The results of the study guide biologists in selecting a visualization tool based on the type of their microarray data, visualization designers on the key role of user interaction techniques, and evaluators on a new approach for evaluating the effectiveness of visualizations for providing insight. Though we used the method to analyze bioinformatics visualizations, it can be applied to other domains.

Cigarette smoke worsens lung inflammation and impairs resolution of influenza infection in mice

BackgroundCigarette smoke has both pro-inflammatory and immunosuppressive effects. Both active and passive cigarette smoke exposure are linked to an increased incidence and severity of respiratory virus infections, but underlying mechanisms are not well defined. We hypothesized, based on prior gene expression profiling studies, that upregulation of pro-inflammatory mediators by short term smoke exposure would be protective against a subsequent influenza infection.MethodsBALB/c mice were subjected to whole body smoke exposure with 9 cigarettes/day for 4 days. Mice were then infected with influenza A (H3N1, Mem71 strain), and analyzed 3 and 10 days later (d3, d10). These time points are the peak and resolution (respectively) of influenza infection.ResultsInflammatory cell influx into the bronchoalveolar lavage (BALF), inflammatory mediators, proteases, histopathology, viral titres and T lymphocyte profiles were analyzed. Compared to smoke or influenza alone, mice exposed to smoke and then influenza had more macrophages, neutrophils and total lymphocytes in BALF at d3, more macrophages in BALF at d10, lower net gelatinase activity and increased activity of tissue inhibitor of metalloprotease-1 in BALF at d3, altered profiles of key cytokines and CD4+ and CD8+ T lymphocytes, worse lung pathology and more virus-specific, activated CD8+ T lymphocytes in BALF. Mice smoke exposed before influenza infection had close to 10-fold higher lung virus titres at d3 than influenza alone mice, although all mice had cleared virus by d10, regardless of smoke exposure. Smoke exposure caused temporary weight loss and when smoking ceased after viral infection, smoke and influenza mice regained significantly less weight than smoke alone mice.ConclusionSmoke induced inflammation does not protect against influenza infection.In most respects, smoke exposure worsened the host response to influenza. This animal model may be useful in studying how smoke worsens respiratory viral infections.

An Insight-Based Longitudinal Study of Visual Analytics

Visualization tools are typically evaluated in controlled studies that observe the short-term usage of these tools by participants on preselected data sets and benchmark tasks. Though such studies provide useful suggestions, they miss the long-term usage of the tools. A longitudinal study of a bioinformatics data set analysis is reported here. The main focus of this work is to capture the entire analysts process that an analyst goes through from a raw data set to the insights sought from the data. The study provides interesting observations about the use of visual representations and interaction mechanisms provided by the tools, and also about the process of insight generation in general. This deepens our understanding of visual analytics, guides visualization developers in creating more effective visualization tools in terms of user requirements, and guides evaluators in designing future studies that are more representative of insights sought by users from their data sets

An Evaluation of Microarray Visualization Tools for Biological Insight

High-throughput experiments such as gene expression microarrays in the life sciences result in large datasets. In response, a wide variety of visualization tools have been created to facilitate data analysis. Biologists often face a dilemma in choosing the best tool for their situation. The tool that works best for one biologist may not work well for another due to differences in the type of insight they seek from their data. A primary purpose of a visualization tool is to provide domain-relevant insight into the data. Ideally, any user wants maximum information in the least possible time. In this paper we identify several distinct characteristics of insight that enable us to recognize and quantify it. Based on this, we empirically evaluate five popular microarray visualization tools. Our conclusions can guide biologists in selecting the best tool for their data, and computer scientists in developing and evaluating visualizations

Epstein-Barr virus: a paradigm for persistent infection – for real and in virtual reality

The really interesting thing about herpesviruses is that they can establish lifelong persistant infections in immunocompetent hosts. At first glance, they would seem to have very different ways of doing this. Here we will use as a model our current understanding of how the human herpesvirus Epstein-Barr virus establishes and maintains such an infection. We apply information from a wide range of sources including laboratory experimentation, clinical observation, animal models and a new computer simulation. We propose that the detailed mechanisms for establishing infection are dependent on the virus and tissues involved, but the strategy is the same - to persist in a long-lived cell type where the virus is invisible to the immune system and nonpathogenic.

Simulating Epstein-Barr virus infection with C-ImmSim

MOTIVATION Epstein-Barr virus (EBV) infects greater than 90% of humans benignly for life but can be associated with tumors. It is a uniquely human pathogen that is amenable to quantitative analysis; however, there is no applicable animal model. Computer models may provide a virtual environment to perform experiments not possible in human volunteers. RESULTS We report the application of a relatively simple stochastic cellular automaton (C-ImmSim) to the modeling of EBV infection. Infected B-cell dynamics in the acute and chronic phases of infection correspond well to clinical data including the establishment of a long term persistent infection (up to 10 years) that is absolutely dependent on access of latently infected B cells to the peripheral pool where they are not subject to immunosurveillance. In the absence of this compartment the infection is cleared. AVAILABILITY The latest version 6 of C-ImmSim is available under the GNU General Public License and is downloadable from www.iac.cnr.it/~filippo/cimmsim.html

A multifactorial role for P. falciparum malaria in endemic Burkitt's lymphoma pathogenesis.

Endemic Burkitts lymphoma (eBL) arises from the germinal center (GC). It is a common tumor of young children in tropical Africa and its occurrence is closely linked geographically with the incidence of P. falciparum malaria. This association was noted more than 50 years ago. Since then we have learned that eBL contains the oncogenic herpes virus Epstein-Barr virus (EBV) and a defining translocation that activates the c-myc oncogene. However the link to malaria has never been explained. Here we provide evidence for a mechanism arising in the GC to explain this association. Accumulated evidence suggests that eBL arises in the GC when deregulated expression of AID (Activation-induced cytidine deaminase) causes a c-myc translocation in a cell latently infected with Epstein-Barr virus (EBV). Here we show that P. falciparum targets GC B cells via multiple pathways to increase the risk of eBL. 1. It causes deregulated expression of AID, thereby increasing the risk of a c-myc translocation. 2. It increases the number of B cells transiting the GC. 3. It dramatically increases the frequency of these cells that are infected with EBV and therefore protected from c-myc induced apoptosis. We propose that these activities combine synergistically to dramatically increase the incidence of eBL in individuals infected with malaria.

A virtual look at Epstein-Barr virus infection: simulation mechanism.

Epstein-Barr virus (EBV) is an important human pathogen that establishes a life-long persistent infection and for which no precise animal model exists. In this paper, we describe in detail an agent-based model and computer simulation of EBV infection. Agents representing EBV and sets of B and T lymphocytes move and interact on a three-dimensional grid approximating Waldeyers ring, together with abstract compartments for lymph and blood. The simulation allows us to explore the development and resolution of virtual infections in a manner not possible in actual human experiments. Specifically, we identify parameters capable of inducing clearance, persistent infection, or death.

PathSim visualizer: an Information-Rich Virtual Environment framework for systems biology

Increasingly, biology researchers and medical practitioners are using computational tools to model and analyze dynamic systems across scales from the macro to the cellular to the biochemical level. We are using Information-Rich Virtual Environments (IRVEs) to display the results of biological simulations, and to allow users to interact with those simulations. While simulation architectures, algorithms, and processing power have enjoyed continuous optimization to date, the user interfaces to these applications have not. The problems of designing such IRVE interfaces arise from the requirement that a variety of spatial and abstract information must be integrated into one coherent experience for the user. This paper explores the design and development issues encountered in our implementation of a bioinformatics application, PathSim (Pathogen Simulation). Specifically, we describe the information and interaction issues in building a front-end tool to visually analyze the results of an agent-based immunology simulation. Finally, we present custom scenegraph objects and consider candidate functionality for future standards components.

A comparison of benchmark task and insight evaluation methods for information visualization

This study compares two different empirical research methods for evaluating information visualizations: the traditional benchmark-task method and the insight method. The methods are compared using criteria such as the conclusions about the visualization designs provided by each method, the time participants spent during the study, the time and effort required to analyse the resulting empirical data, and the effect of individual differences between participants on the results. The study compares three graph visualization alternatives that associate bioinformatics microarray time series data to pathway graph vertices in order to investigate the effect of different visual grouping structures in visualization designs that integrate multiple data types. It is confirmed that visual grouping should match task structure, but interactive grouping proves to be a well-rounded alternative. Overall, the results validate the insight method’s ability to confirm results of the task method, but also show advantages of the insight method to illuminate additional types of tasks. Efficiency and insight frequently correlate, but important distinctions are found. Categories: H.5.2 [Information Interfaces and Presentation]: User Interfaces – evaluation/methodology.