GeVaDSs – decision support system for novel Genetic Vaccine development process
© Blazewicz et al.; licensee BioMed Central Ltd. 2012
Received: 10 October 2011
Accepted: 11 April 2012
Published: 10 May 2012
The lack of a uniform way for qualitative and quantitative evaluation of vaccine candidates under development led us to set up a standardized scheme for vaccine efficacy and safety evaluation. We developed and implemented molecular and immunology methods, and designed support tools for immunization data storage and analyses. Such collection can create a unique opportunity for immunologists to analyse data delivered from their laboratories.
We designed and implemented GeVaDSs (Genetic Vaccine Decision Support system) an interactive system for efficient storage, integration, retrieval and representation of data. Moreover, GeVaDSs allows for relevant association and interpretation of data, and thus for knowledge-based generation of testable hypotheses of vaccine responses.
GeVaDSs has been tested by several laboratories in Europe, and proved its usefulness in vaccine analysis. Case study of its application is presented in the additional files. The system is available at: http://gevads.cs.put.poznan.pl/preview/(login: viewer, password: password).
Immune response & vaccination principles
Vaccines are effective tools to prevent infectious diseases in humans or animals [1, 2]. A vaccine typically contains an antigen mixture that stimulates a body’s immune system to recognize the agent as foreign, inducing specific immune responses and memory for long-term protection against the disease-causing microorganisms. Vaccines can be prophylactic (to prevent a future infection) or therapeutic (e.g. vaccines against cancer). Following vaccination, both innate and adaptive immune system components synergize to elicit an immune response. Antigen-presenting cells - notably dendritic cells - take up antigens and traffic to the draining lymph nodes where they present processed antigens to naive CD4+ and CD8+ T lymphocytes. Naive T-cells are stimulated to proliferate and differentiate into effector and memory T-cells. Activated effector and memory CD4+ T-cells provide help to B-cells to mount antibody responses, and help naive CD8+ T-cells to enhance their clonal expansion and differentiation into cytotoxic CD8+ T lymphocytes (CTL). The quality of the vaccine-induced immune response depends on several factors, e.g. antigen nature, route of administration, antigen presentation, vaccine preparation adjuvants, or timing between challenges.
Vaccine development & evaluation
Usually, the effectiveness of vaccination is ascertained when vaccinated individuals exposed to infection are protected. To be effective, a vaccine should trigger efficient activation of antigen-presenting cells to initiate antigen processing and presentation to T-cells, as well as activation of T and B-cells (production of antibodies, generation of memory B-cells, memory CTL and memory CD4+ T-cells). Several immune parameters can be measured to monitor the immune response, including antigen-specific and neutralizing antibodies, antigen-specific T-cell expansion, CTL activity and cytokine production. Dendritic cell state of activation can be assessed by transcriptome analysis. A central goal of vaccine research is to identify whether an early vaccine-induced immune response is predictive of later protection [3, 4]. An immune response can be used for guiding vaccine development, for predicting vaccine efficacy in different settings, and for setting vaccination policies and regulations.
The new generations of recombinant vaccines comprise peptides, DNA vaccines, viral or bacterial delivery system and Virus Like Particles. Although enormous progress has been made in the process of vaccine development, there is still a need to standardize the qualitative and quantitative evaluation of new vaccines, to develop a platform of novel recombinant genetic vaccines using genomic and proteomic information and to create interactive systems to store all types of experimental data together (such as immune response parameters stated above) with statistical and analytical tools to support scientists during their experiment. The above goals were achieved by the CompuVac consortium  established within a project funded under FP6 EC programme. As one of the key deliverable of this project, a new web platform called GeVaDSs (Genetic Vaccine Decision Support system) has been designed and implemented. GeVaDSs is accessible for scientific community. GeVaDSs allows one to (i) find the description of stored experiments, (ii) create new experiments and store experimental data, and (iii) visualize and cross-compare the results of theses experiments. The construction and the development of the GeVaDSs is a first trial to standardize and describe the whole process of vaccine development, making it a really challenging aspect of the CompuVac project.
Although it is difficult to find efficient system for vaccine analysis, it is worth to mention results of VIOLIN initiative (Vaccine Investigation and Online Information Network) [6, 7]. It is a system for storing and analyzing published vaccine data. It has been developed for data mining, curation and analysis of results for commercially available vaccines, as well as vaccine candidates on the early stage of development. The main difference between VIOLIN and the one presented in GeVaDSs is the level of data collected in the system and the analysis possible. The first system gives a researcher an opportunity to gather data concerning a particular vaccine. The second one is useful at the level of laboratory tests of those vaccine. GeVaDSs was developed to store raw experimental results and presents them in a user friendly form. It allows for a comparison of the organism response to a vaccination at different levels like T-cell and B-cell. Furthermore, the influence of the vaccine on the selected gene expression can be determined in the Molecular Signature module and the efficacy of the immunization protocol used, is analyzed in the Vector Challenge module.
The system was implemented using the Adobe Flex 3 technology and Action Script 3 language on the client side (interface, logic mechanisms), whereas the server side interface was implemented using mainly the PHP script language, in addition to other languages like Perl, R and Java. All data processed by the GeVaDS system are stored in the PostgreSQL relational database.
Structure of the system
The vaccine-related modules are: T-cell, B-cell, Molecular Signature, Vector Challenge, Evaluation and Dictionaries (the latter is shared by all other modules and is not shown in Figure 1). Their brief descriptions are presented below. The support modules are: input for introducing data, output for viewing the results, Quality Check - for checking the quality of introduced data, reports which allows the user to generate a report for requested data as a pdf file, and the system access and privileges. All of the support modules, excluding the last one, are designed separately for each of the vaccine-related modules. For example, GeVaDSs consists of T-cell input module, B-cell input module etc. Obviously, input modules are very similar.
As mentioned above, Dictionaries is a general module, which is shared by all modules to unify terms and definitions used in experiments definition. It defines basic terms like species, animal models, antigens... used across the whole system. However, some parts are crucial to understand the idea of GeVaDSs, and therefore are presented below in more details.
Hierarchy of vaccine vectors
Organization of experiments
Experiment is the basic assessment unit in the system. It describes a vaccination experiment conducted in the laboratory, and stores data such as: information about experiment creator, date, default parameters of the experiment (default animal model, default sex, default age...) which would be inherited by the experimental group unless otherwise specified. Default parameters are thus helpful when many experimental groups are similar to each other only with slight differences. Each experiment introduced into the system involves several experimental groups. An experimental group consist of a set of uniform individuals (same age, same sex, same animal model...) for which the immunization followed a given protocol.
There are five types of experimental group. Each experimental group can be one of the following types:
· experimental vaccine - a group of individuals being immunized with an experimental vaccine to be tested.
· naive/control - a group immunized with an empty vector, without antigen.
· negative control - no immunization.
· Internal Standard - a group being immunized using a predetermined standard immunization protocol based on the selected internal standards, which constitutes a benchmark for comparison of different experiments (e.g. carried out in different laboratories).
System access & privileges
The usage of the GeVaDS system requires authorization. This mechanism not only protects the system against accidental usage or misbehavior, but also ensures safety of data stored in the system. Authorized users have access to modules depending on the role assigned by the system administrator (e.g. admin, lab manager, etc). Privileges defined in the system are divided into two groups: those connected with the access to particular modules or with performing some actions, and those connected with the user data. Privileges for the first group could be assigned by the system administrator (they are called Roles). The second type of privileges are managed by users themselves. Each user has a full access to his/her own data (as the owner of the data), a read-only access to public data and access to data shared by other users. The owners can make their data publicly accessible or they can give a custom access to other users. However, by default, the lab manager (a special role) of a given laboratory has full access to all data of users belonging to his/her laboratory.
Quality Control module
Quality Control module is responsible for checking the quality and correctness of the data introduced by the user as well as the composition of experiments. The quality control process is based on a set of quality rules. Each rule is assigned a unique number, information about the data module to which the rule should be applied (T-cell, B-cell, Molecular Signature etc.) and the level of application (experiment, experimental group, individual or data value). Depending on the level of application, there are rules responsible for checking the definition of experiments or experimental groups (i.e.: Have the laboratory & date of experiment been provided? Have experimental control groups been created? Is each experimental group large enough for statistical analysis?...), as well as controlling every single value of all attributes (i.e.: Are interferon gamma production values provided as percentages? Are the naive experimental group results in acceptable ranges?...). Reference values of attributes defining acceptable ranges have been defined by immunology experts.
The procedure of quality control of an experiment (or an experimental group) should be launched by the user before data analysis. During this procedure, in case a rule is broken the system generates an error message. At the end of the procedure a report listing all messages is produced. Error messages belong to the followings categories:
· minor error
· serious error
· fatal error
After the quality control procedure, each experiment, as well as experimental group, receives a quality status. Five different quality statuses have been defined:
· Not Checked (NC) - the default status before launching the quality control procedure. Quality status is reset to NC after any revision.
· Rejected (R) - the data for experimental group or experiment have some fatal errors and therefore the data need to be corrected before proceeding.
· Pending (P) - the data for experimental group or experiment have some serious errors; the user may decide to manually enforce the quality status from Pending to Rejected or User.
· User (U) - the data for experimental group or experiment have been manually validated by the user (see Pending).
· Auto (A) - the data for experimental group or experiment have no fatal nor serious errors and have been automatically validated by the system.
Each experiment may also have one of two quality flags:
· E-ready - means that the experiment has reliable data and involves correct experimental groups that can be analyzed. The E-ready flag is received when at least one Naive and one Internal Standard or Experimental Vaccine groups have the Auto or User quality status.
· EC-ready - means that data stored for this experiment can be analyzed and compared against other experiments. The EC-ready flag is received when at least one Naive, one Internal Standard, and one Experimental Vaccine groups have the Auto or User quality status.
T-cell module analyzes the response mediated by T lymphocytes over time. This response is described by the following parameters: expansion of vaccine antigen-specific T-cells (%), memory phenotype (%), CTL activation (% of intracellular interferon gamma positive CD8+T-cells) and cytotoxicity (% of specific lysis).
T-cell response values can be introduced in two ways: manually, by introducing values using the T-cell input submodule or automatically, by generating and uploading a standardized template (a Microsoft Excel file) that can be filled by the user by copy-paste from his/her worksheet and then uploaded back into the system. After checking for consistency, data is automatically inserted and quality-controlled.
These alternative input methods raise the problem of a possible conflict between automatic uploading of a template and manual edition of the experiment after template generation. This problem has been circumvented by stamping both the template and experiment at the time of template generation, the experiment stamp being modified during any manual edition. Therefore, in the latter case, an older template version will be rejected for uploading.
Once data have been properly uploaded and quality-controlled, they can be analyzed with T-cell response submodules:
B-cell response module analyzes the response mediated by B lymphocytes in terms of antibody production and antibody neutralization activity. The results are presented as sigmoid curves obtained after linear regression for serum titration . Since calculating the sigmoid is computationally time consuming, the system calculates the sigmoid once on first request. Linear regression results are stored in the database.
For each experiment, the following parameters should be defined:
· serum descriptions (typically marked as S0, S1...) - which are defined for different days of measurement,
· serum dilutions (marked as 1/20, 1/100...) - which are defined for different dilutions tested.
The structure of the B-cell response module is very similar to that of the T-cell response module. The main difference concerns the data insertion requirement since the definition of the experiment is a prerequisite step before template creation, as well as the definition of serum descriptions and dilutions. Once experiment is properly described, the B-cell template is found very convenient with regard to the much higher amount of data to be provided for titration experiments.
The sigmoid titration curve is expressed as follows:
· with the sigmoid formula (Eq. 1) ,
· with a table with the sigmoid parameters values (not shown),
· top - is a maximal curve asymptote,
· bottom - is a minimal curve asymptote,
· HillSlope- denotes the steepness of the curve,
· EC 50 - denotes the dilution that produces the neutralization halfway between the bottom and top response levels.
Again, several output submodules can be distinguished in the B-cell module.
Experimental group submodule shows the sigmoid for all individuals assigned to the selected group, and the average sigmoid (determined by a linear regression for all data points). The results also contain the set of charts with the neutralization per dilutions for Ag-pp and CTR-pp together with the average values of dilutions.
The outcome provides normalized sigmoids for all selected experimental groups, quality charts and data tables.
Molecular Signature module
The main purpose of the Molecular Signature module is to extract a gene subset, the expression of which varies after immunization. Such gene signatures could then be used as markers of vaccine efficacy. Molecular Signature module characterizes the change of gene expression of vaccinated individuals against that of control individuals. Input data for this module are results of microarray experiments which are introduced into the system. The organization of the experiment data is similar to that in the other modules. Hence, a user defines an experiment, experimental groups, and individuals within each experimental group. For each individual, microarray results can be introduced for various organ and cell types. Currently, the system is able to analyze Illumina, CodeLink & Affymetrix microarray data. Each experiment, as well as experimental group, has a quality status which defines the quality of the data (see Quality Control module section).
Gene expression analysis can be restricted to predefined gene lists of interest, created and managed by users.
Vector Challenge module
Vector Challenge module measures the efficacy of immunization protocols. It stores survival, or protection, information of vaccinated individuals challenged with live virus (against which the vaccine, if efficient, is expected to protect the individual).
Again, Vector Challenge result parameter values can be inserted by generating an MS Excel template file after defining necessary information about the experiment, experimental groups and number of individuals within each group. The template can be filled by the user by copy-pasting from his/her worksheet and then uploading back to the system. After checking for potential inconsistencies, uploaded data will be inserted automatically. In case of errors, a report is generated. The mechanism of template generations and template uploading is similar to that in T-cell and B-cell modules.
The following data analysis submodules have been developed:
· Individual submodule, which allows the user to analyze single individuals. Data can be presented in a tabular way and as a chart.
· Experimental group submodule, which allows the user to view an outcome of challenge results for a selected experimental group presenting the average values for individuals. Results are presented in a chart as well as a data table (not shown).
The idea and implementation of the Evaluation module is a unique solution in the current state of the art of immunological experiments. The idea that lies behind this module is to assess experiments using multicriteria analysis. Evaluation module provides an overall evaluation of investigated vaccines and vectors. The evaluation process establishes effectiveness of a vaccine, based on the results received for the monitored parameters for all types of experiments (T-cell response, B-cell response, Molecular Signature).
Result and discussion
An end-user demonstration is provided as Additional file 1 in order to give a practical tutorial for GeVaDSs main features. The data used for this demonstration example are based on current experimental data produced during the course of the CompuVac project; they are available in GeVaDSs for demonstration purpose so that the corresponding analyses can be reproduced. This simulated example of GeVaDSs illustrates how to test the potential strength of a newly isolated virus (i.e. PEV) as a new vaccine platform by: (i) evaluating the potential of this vaccine platform candidate and (ii) comparing it to other available vector platforms already tested in GeVaDSs. First, the three vector constructs to be tested (PEVdelta, PEVdelta-gp33, PEVdelta-VSVG) are defined in GeVaDSs (slides 3-7). Mice are then immunized with these constructs or CompuVac reference vector (Internal Standard) (slides 8-12). The corresponding T-cell & B-cell immune responses as well as transcriptome data are collected and automatically entered in GeVaDSs database (slides 13-18). Finally, the appropriate analysis reports are generated in order to evaluate PEV vector performance (slides 19-24). As a result PEV is positively evaluated as an efficient vector platform (slide 25) (see Additional file 1). The diagram illustrating the phases of data processing is presented in the chart - see Additional file 2.
GeVaDSs, developed by the CompuVac consortium, offers a unique solution for improved development of novel Vaccines by the ability to analyse and compare experimental data from different laboratories. The interactive database contains standardized data related to a series of recombinant vaccines (>100) covering most categories of currently developed vaccine platforms (peptide, DNA, viral-based, bacterial-based and VLPs). This data set may serve as a core reference for vaccine developers. The use of the CompuVac Internal standards allows for newly generated data to be added and reliably compared with the core database. During the CompuVac project timeframe, GeVaDSs has already allowed users to conduct comparisons between different vaccine types and initiated novel vaccine design and vaccination regimens. For example, GeVaDSs identified an adenovirus/DNA prime-boost immunization scheme as more efficient than the conventional DNA/adeno prime-boost scheme for vaccination against hepatitis C virus . Besides monitoring of T- and B-cell immune responses, vaccine efficacy and safety profiles were also analyzed by transcriptome studies which allowed the system to generate molecular signatures of different vectors, and their clustering in vector classes. Detailed analyses of these signatures is underway to decipher the underlying networks/pathways correlating with vector activity (manuscript in preparation). We believe that GeVaDSs, whose database is constantly growing, should contribute to rationalizing and speeding up vaccine development.
GeVaDSs has afforded integration of biological and bioinformatics components of immunological experiments. Together with other similar initiatives [7, 10], the system provides valuable evidence of the progress in this area and delivers user friendly interface and tools for future analysis of experiments. Recent successful systems biology approaches in the field of immune/vaccine response analysis advocate for larger initiatives and development of standardized databases [11–13]. The flexibility of GeVaDSs makes it possible to adapt the system in a wider field of usage than that specific to CompuVac. This might be of value for acquiring and analyzing datasets associated with translational research investigation in medicine, where multiple parameters are considered with increasing potential for integration of transcriptome, proteome and other datasets, requiring sophisticated bioinformatics-driven analysis, but overlap with more conventional clinical measures. In particular, GeVaDSs is able to assess quantitative measures as employed in clinical trial assays for validation concerning reproducibility, inter-assay and inter-lab variation. GeVaDSs therefore should contribute to shifting traditional and empirical vaccine development to systems vaccinology .
Availability and requirements
· Project name: GeVaDSs
· Project home page: http://gevads.cs.put.poznan.pl/,
· Trial version open to the public with read-only rights: http://gevads.cs.put.poznan.pl/preview/(login: viewer, password: password).
· Operating system: Linux for server side
· Programming language: client side - Action Script 3.0 and Flex, server side - PHP, c/c++ with GNU Scientific Library, Perl and R.
· Other requirements: client side - web browser with Flash Player v. 9.0 or newer, server side - PostgreSQL, Apache 2.0
· Any restrictions to use by non-academics: none
The full version of GeVaDSs is publicly available and any researcher can use the system for free after registration. However, the source code of the system is not publicly available and there is no distributable version because of the complexity of implemented modules. The system consists of many connected modules written in many programming languages and installing it is very complicated. In special cases, we are ready to consider an installation of the system for an interested user on his dedicated server.
DK was the main coordinator of the COMPUVAC FP6 EU project. DK and AS proposed the idea of GeVaDS, whereas JB was a head of the group which was responsible for its creation and guided the final preparation of the paper. AS specified system requirements from biological point of view. PL was a manager of the GeVaDSs’ programmers group. MB, PK and PW designed the system architecture and implemented the whole system. AS, WC were responsible for data curation. WC, AS, PL, MB, PK, PW and others participated in many discussions about the outlook of GeVaDSs modules, what was necessary in this multidisciplinary project. All authors were involved in writing the manuscript and all of them read and approved this version.
Sponsored by European Committee under contract No. LSHB-CT-2004-005246 and Polish Ministry of Science and Higher Education under Grant No.: 136/E-362/6.PR UE/DIE 206/2005-2008 and partially supported by grant NO DEC-2011/01/B/ST6/07021 from the National Science Centre, Poland. Anna Tchassovskikh is acknowledged for her input in the design of the Quality Check module. Nicolas Derian is acknowledged for helpful contribution in beta-testing, database curation, quality control and microarray submodules specification, critical reading of the manuscript.
- Taylor K, Nguyen A, Stephenne J: The need for new vaccines. Vaccine 2009, 27(Suppl 6):G3—G8.PubMedGoogle Scholar
- Plotkin SA: Vaccines: the fourth century. Clin Vaccine Immunol 2009, 16(12):1709–1719. 10.1128/CVI.00290-09PubMed CentralView ArticlePubMedGoogle Scholar
- Oberg AL, Kennedy RB, Li P, Ovsyannikova IG, Poland GA: Systems biology approaches to new vaccine development. Curr Opin Immunol 2011, 23(3):436–443. . [http://www.sciencedirect.com/science/article/pii/S0952791511000409]. 10.1016/j.coi.2011.04.005PubMed CentralView ArticlePubMedGoogle Scholar
- Bellier B, Six A, Thomas-Vaslin V, Klatzmann DSpringer Verlag, Heidelberg New York (in press) Springer Verlag, Heidelberg New York (in press)
- Compuvac Project . [http://compuvac.cs.put.poznan.pl].
- Xiang Z, Todd T, Ku K, Kovacic B, Larson C, Chen F, Hodges A, Tian Y, Olenzek E, Zhao B, Colby L, Rush H, Gilsdorf J, Jourdian G, He Y: VIOLIN: Vaccine Investigation and Online Information Network. Nucleic Acids Res 2008, 36: D923-D928.PubMed CentralView ArticlePubMedGoogle Scholar
- Vaccine Investigation and Online Information Network 2011.
- Motulsky H, Christopoulos A: Models to Biological Data Using Linear and Nonlinear Regression. 2004.Google Scholar
- Garrone P, Fluckiger AC, Mangeot PE, Gauthier E, Dupeyrot-Lacas P, Mancip J, Cangialosi A, Du Chene I, LeGrand R, Mangeot I, Lavillette D, Bellier B, Cosset FL, Tangy F, Klatzmann D, Dalba C: Sci Translational Med. 2011, 3(94):94ra71.http://stm.sciencemag.org/content/3/94/94ra71.abstract.Google Scholar
- Kholer A, Puzone R, Seiden P, Celada F: A systematic approach to vaccine complexity using an automaton model of the cellular and humoral immune system. I. Viral characteristics and polarized responses. Vaccine 2000, 19(7–8):862–76. 10.1016/S0264-410X(00)00225-5View ArticleGoogle Scholar
- Chaussabel D, Quinn C, Shen J, Patel P, Glaser C, Baldwin N, Stichweh D, Blankenship D, Li L, Munagala I, Bennett L, Allantaz F, Mejias A, Ardura M, Kaizer E, Monnet L, Allman W, Randall H, Johnson D, Lanier A, Punaro M, Wittkowski KM, White P, Fay J, Klintmalm G, Ramilo O, Palucka AK, Banchereau J, Pascual V: A modular analysis framework for blood genomics studies: application to systemic Lupus Erythematosus. Immunity 2008, 29: 150–164. . [http://www.sciencedirect.com/science/article/pii/S1074761308002835]. 10.1016/j.immuni.2008.05.012PubMed CentralView ArticlePubMedGoogle Scholar
- Querec TD, Akondy RS, Lee EK, Cao W, Nakaya HI, Teuwen D, Pirani A, Gernert K, Deng J, Marzolf B, Kennedy K, Wu H, Bennouna S, Oluoch H, Miller J, Vencio RZ, Mulligan M, Aderem A, Ahmed R, Pulendran B: Systems biology approach predicts immunogenicity of the yellow fever vaccine in humans. Nat Immunol 2008, 10: 116–125.PubMed CentralView ArticlePubMedGoogle Scholar
- Park AW, Daly JM, Lewis NS, Smith DJ, Wood JLN, Grenfell BT: Quantifying the impact of immune escape on transmission dynamics of influenza. Science 2009, 326(5953):726–728. . [http://www.sciencemag.org/content/326/5953/726.abstract]. 10.1126/science.1175980PubMed CentralView ArticlePubMedGoogle Scholar
- GeVaDSs Manual . [http://compuvac.cs.put.poznan.pl/resources/userfiles/files/Gevads-manual.pdf].
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.