Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Contents lists available at ScienceDirect
Computers in Biology and Medicine journal homepage: www.elsevier.com/locate/cbm
BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data Yuri Ferretti a,b, Newton Shydeo Brandão Miyoshi a, Wilson Araújo Silva Jrc,d, Joaquim Cezar Felipe a,b,c,n a
Inter-institutional Post-graduation Program on Bioinformatics, University of São Paulo, Brazil Department of Computing and Mathematics, Faculty of Philosophy Sciences and Languages of Ribeirão Preto, University of São Paulo, Brazil c Center for Integrative Systems Biology - CISBi, NAP/USP, University of São Paulo, Brazil d Department of Genetics, School of Medicine of Ribeirão Preto, University of São Paulo, Brazil b
art ic l e i nf o
a b s t r a c t
Article history: Received 30 April 2014 Accepted 4 April 2015
Background: Researchers of translational medicine face numerous challenges in attempting to bring research results to the bedside. This field of research covers a wide range of resources, including blood and tissue samples, which are processed for isolation of RNA and DNA to study cancer omics data (genomics, proteomics and metabolomics). Clinical information about patients' habits, family history, physical examinations, remissions, etc., is also important to underpin studies aimed at identifying patterns that lead to the development of cancer and to its successful treatment. Purpose: Development of a web-based computer system—BioBankWarden—to manage, consolidate and integrate these diversified data, enabling cancer research groups to retrieve and analyze clinical and biomolecular data within an integrative environment. The system has a three-tier architecture comprising database, logic and user-interface layers. Results: The system's integrated database and user-friendly interface allow for the control of patient records, biomaterial storage, research groups, research projects, users and biomaterial exchange. Conclusions: BioBankWarden can be used to store and retrieve specific information from different clinical fields linked to biomaterials collected from patients, providing the functionalities required to support translational research in the field of cancer. & 2015 Elsevier Ltd. All rights reserved.
Keywords: Translational medicine Biological databases Oncogenomics Biomaterial management Clinical data management
1. Introduction Human genome mapping and the emergence of new sequencing techniques have motivated cancer researchers to focus a great deal of attention on understanding the mechanisms that contribute to the development of oncology diseases. This includes translational medicine, which combines basic, social and political sciences to optimize patient care and preventive practices and to identify behavioral and environmental factors that can trigger mutations in the human genetic code, which can give rise to a variety of tumors [1]. Translational medicine involves the application of scientific research results, particularly those of omics techniques, to improve health and treatment processes [2], aiming to bridge the gap between research and clinical practice.
n Corresponding author at: Av Bandeirantes, 3900, Ribeirão Preto, SP, 14040-901, Brazil. E-mail addresses: [email protected] (Y. Ferretti), [email protected] (N.S.B. Miyoshi), [email protected] (W.A. Silva Jr), [email protected] (J.C. Felipe).
Scientists usually investigate abnormal levels of gene expression that could be related to specific cancer diseases in order to identify the gene responsible for the disorder. This requires the collection of tissue and blood samples from patients to isolate DNA, RNA and proteins, which can reveal genetic mutations. A concomitant analysis of patients' personal and clinical data, such as family history, smoking and drinking habits, race, medical history, etc., can also be performed to identify cancer-related factors. Finally, the data obtained from these different sources can be analyzed jointly to determine which environmental or behavioral factors play meaningful roles in cancer genesis. The purpose of this study was to develop a web-based computer system—called BioBankWarden (BBW)—which enables researchers to store, retrieve and integrate biomolecular and clinical data on biomaterials collected from patients, as recommended by Brisk research-oriented storage kit [3]. BBW supports the management of biomaterial data, such as requisitions and submissions of aliquots, to help researchers to determine the requirements for their studies. Another important point is to manage user permissions to access data from specific research groups, avoiding ethical issues. To facilitate the control and access of patient data from different clinical
http://dx.doi.org/10.1016/j.compbiomed.2015.04.008 0010-4825/& 2015 Elsevier Ltd. All rights reserved.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
2
areas, BBW provides different data sets for each clinical area, according to the user's needs. The Bank of Central Nervous System Tumors, which emerged as a subproject of the Clinical Genome Project, an initiative of the Department of Neurosurgery of Ribeirão Preto Medical School (RPMS) – University of São Paulo (USP) – Brazil, initially contained samples of only one type of nervous system tumor. This initial collection of biological material was later expanded to encompass all the tumors of the central nervous system, as well as other diseases treated by the neurosurgical team of the Teaching Hospital of RPMS, i.e., urology, gastric surgery, proctology and gynecology. The Biobank Nucleus for Research Support, which was created in 2011 through a collaboration of several researchers at the Translational Laboratory, also contains a large number of samples. In 2012, CISBi (Center for Integrative Systems Biology) was created to provide computational tools for the analysis, integration and interpretation of biological data and to encourage the formulation of new concepts involving relevant biological issues, such as dynamic systems theories, analysis of molecular noise, and statistics applied to genetic networks. The development of the BBW system is the result of the Biobank and CISBi merger. BBW will support the storage, management and integration of data from all the projects related to these initiatives. The researchers who participate in the projects contribute by detailing the system requirements and evaluating the prototypes. Several computational platforms have been designed to store and query biomolecular and clinical data. Some examples are: SlimPrim (Scientific Laboratory Information Management – Patient-care Research Information Management) [4]; STRIDE (Stanford Translational Research Integrated Database Environment) [5]; and I2B2 (Informatics for Integrating Biology and the Bedside) [6]. Miyoshi et al. [7] implemented the IPTrans (Integrative Platform for Translational Research), a framework for integrating biomolecular and clinical data based on the use of a common reference ontology. The clinical database of this platform is based on the Entity-AttributeValue data structure. However, this platform does not provide tools for controlling the storage and distribution of biomaterial samples or for discriminating specific clinical data according to different clinical areas (digestive tract, female reproductive tract, etc.) or for arranging data and users according to research groups and projects. IPTrans framework supports the integration of heterogeneous data sources containing personal, demographic and clinical information about patients. It has also a database to store results of biomolecular procedures, such as levels of genes expression, performed on the same patients' genetic material. In addition, an analytical module is under development, which will enable the user to perform analyses by merging clinical data and biomolecular results. As part of this environment, BBW will be a clinical data source of IPTrans. The results of biomolecular procedures performed on BBW biomaterials will be stored in the biomolecular database of IPTrans and then the analytical module will be able to perform the clinical/omics integrative analysis. Thus, the main benefit we expect from BBW is that it can act as a clinical data source for a broader environment, namely the IPTrans, thereby promoting the desired linking with biomolecular data, thus enabling the user to accomplish integrative analyses. Other major contribution of the system's implementation is its capability to tightly control storage and usage of biological materials. Furthermore, BBW enables the management of information about patients and biomaterials by assigning them to research projects and/or groups and also defining a user hierarchy of access in such a way to guarantee security and privacy of sensitive data.
2. Similar studies Elliot and Peakman [8] describe the UK Biobank system as a large database composed of biological samples collected from
500,000 patients between 2007 and 2010. The clinical data, which comprise physical and cognitive measurements, were collected by means of questionnaires, and the biological samples consist of urine and blood. All the samples are stored and managed by a proprietary LIMS (Laboratory Information Management System). In Norway, Ronning et al. [9] created a biobank with 380,000 biological samples comprising mainly whole blood, plasma, DNA and urine from pregnant women, their partners and their children. The LIMS manages de-identification, location of the aliquots in freezers, and quality control of the entire process. Roden et al. [10] developed a DNA biobank linked to clinical information from a de-identified electronic medical record (EMR) system. This biobank receives a constant supply of new DNA samples, at a rate of 700–900 samples per week. De-identification and sampling algorithms were developed and integrated into the biobank information system. In Taiwan [11], a biobank was created to collect DNA from a large group of people, aiming to track their health and lifestyle for at least 10 years. The participants will be volunteers recruited throughout the country, in the 30-year age range and of both sexes. The main goal is to track biological data on common chronic diseases such as high blood pressure, diabetes and cancer. Although other research groups have implemented biobank systems, most of these systems focus on specific diseases or conditions, e.g., the Norway Biobank, which focuses on obtaining biological material from pregnant women. Other systems focus on a specific type of biomaterial such as DNA. Another limitation of these biobanks is the fixed size of samples, such as the UK Biobank, which does not accept new biomaterials over time. In addition, patients' clinical and demographic information in these biobanks established a priori, and therefore does not allow for the addition of new clinical information. Unlike the above cited systems, the computational framework and functionalities of BBW, which are described below, appear to be more generic, robust and flexible than the existing solutions. Another relevant feature of BBW that differentiates it from other proposals is its capability to support the management of users, research groups and projects. In contrast to other systems, BBW allows for the creation and association of different projects involving different groups of users, and has a requisition module to manage the output of material based on roles and permissions for exchanges among collaborative groups.
3. Methodology To build BBW, we have been using standard methodologies for system design, development and deployment. Up-to-date technologies and open-source software enables data integration and access of users from different locations, using any web platform. 3.1. Architecture of BioBankWarden BBW is a web-based information system following a clientserver architecture. It is composed by a three-tier architecture: database, logic and user-interface tiers, respectively responsible for storing, processing and presenting the data (Fig. 1). The user-interface tier provides a friendly and intuitive interface that can be accessed from desktop computers, laptops and post-PC devices, such as tablets and smartphones. The database tier is responsible for storing and retrieving all data (clinical and biomaterial) related to the research projects. Finally, the business logic tier comprises four modules: Security module, Biomaterial Module, Data Integration Module and Clinical Module. In these modules all the methods and functions responsible for exchanging information between the database and the user-interface tiers are implemented.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
This tier also performs the validation of data inserted by the user to prevent inconsistencies and controls the user permissions for access, visualization and edition of data. 3.1.1. Security module The security module is responsible for implementing registration, authentication and authorization protocols. The research data stored in BBW demands the creation of many levels of access, restricting or enabling access to the system functionalities. One important assignment of BBW is to allow independent research groups to search biomaterials that could be relevant to their ongoing projects. Any researcher can register to access the system and search for biomaterials of interest and then request these biomaterials. On the other side, hospitals and biomedical centers can also dispose their biomaterials to be managed and accessed through BBW. To provide this flexibility and also guarantee privacy of patient sensitive data we defined a hierarchy of access. This hierarchy allows users to be classified into two distinct roles in the system: users belonging to a disease research group and independent researchers. The users belonging to a disease research group can register new biomaterials and are responsible for the clinical information UI
Independent Researcher
Member Researcher
User interface tier
Security Module Business logic tier Biomaterial Module
Data Integration Module
Clinical Module
Database tier
Database BioBankWarden Fig. 1. Conceptual schema of BBW architecture.
3
about the patients associated to the group. Within a research group there are three types of users: coordinator, medical resident and member. The coordinator can invite new users to take part in the research group, managing group members. The medical resident is responsible for registering patient clinical information and biological samples associated with the patient. Members only have the permission to visualize the clinical data. The other main user, the independent researcher, can ask for the creation of projects and consequently seek and request biomaterials. Within a project there are the following roles: project coordinator, who performs the management of the members of this project; resident, who can make requests of biological samples; and guest, who can view the documents and information from biological samples associated with the projects, but cannot request or edit any information associated to biomaterials. This division into two main groups is also important to guarantee patient privacy. Independent research users are not allowed to access the patient sensitive information, but only those associated to the biomaterial. In addition to these users types there are the curators, responsible for assigning samples to projects and controlling them, and the system administrators, who have full control of the system and can register new research groups and manage clinical surveys. This user hierarchy schema is shown in Fig. 2.
3.1.2. Clinical module BBW is flexible on dealing with various types of biological samples originated from different cancer studies. Each group of patients can have different sets of clinical information available. This information is stored in BBW through Clinical Surveys managed by the Clinical Module. Each disease research group is responsible for providing the clinical information about their patients. The resident physicians that belong to some disease research group is the user responsible for doing the patient clinical data registration using a set of clinical surveys. The surveys are divided into two types: general and research. The general surveys are required for all disease research groups. There are two general surveys representing identification and clinical
Fig. 2. BBW user hierarchy schema.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
4
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
information about the patients: identification survey contains information such as documents, address and contact; and clinical survey contains information such as age, weight, height, family history, use of alcohol, smoking history, etc. The research survey forms vary according to the domain of interest of the disease research groups. Each group can share or create surveys themselves. Currently there are four research surveys: clinical laboratory, anatomic pathology, imaging examination and follow-up. All of them are related to the oncology domain. The clinical laboratory survey represents general information about the tumor, the preoperative, surgery and postoperative stages. The anatomic pathology survey represents more specific information on tumor type, diagnosis and gene expression. The follow-up survey aggregates information about the treatment applied (chemotherapy, radiotherapy) and its output (cure, death, recurrence). The image examination survey allows the addition of images associated to the tumor.
3.2. Technology BBW is built by Web 2.0 technologies, allowing fast prototyping, flexible interface and robust data management. It uses PostgreSQL (version 9.0) [13] as the relational database management system. The logic and user-interface tiers were built using the python object oriented web development framework called Django (version 1.4) [14]. BBW follows the Model-View-Controller (MVC) pattern for software design, enabling clear separation from business logic through declaration of Models, user-interface with the Views and the information flow management in Controllers. To build a clean and ease-to-use interface we have used the library Twitter-bootstrap (version 2.3.2) [15], a front-end web development framework that helps developers to easily and quickly design web pages. The system is deployed with the Apache HTTP Server [16] in Ubuntu Server version 14.04. 3.3. Development strategy
3.1.3. Biomaterial module The Biomaterial Module is responsible for the most important assignment of BBW: the biomaterial management. It allows the system user to control storage, retrieval and access to biomaterial samples. The system accepts several types of samples like blood, serum, tumor margin and others. The volume of a stored sample can be fractioned in smaller amounts, in order to facilitate the storage or distribution among researchers. In the Biomaterial Module the user can also generate processed samples, i.e. DNA, RNA and protein, extracted from the original biomaterial samples. Another requirement is to control how biomaterial samples are sent to researchers that request it. The system allows researchers to request amounts of biomaterial samples, but a coordinator from the owner disease research group must approve each request. It is also possible to filter results when the user searches for stored biomaterials that were extracted from patients with specific characteristics, such as age, weight, smoking history and gender, among others.
The software development life cycle of BBW is divided into three phases: design phase, development phase and deployment phase. In each phase we have adopted a different methodology. In the design phase we have followed the prototyping software development methodology [17]: initial meetings with research groups to analyze the system requirements such as the clinical data to be stored, the users roles for viewing and modifying data and the biomaterial relevant data. Thus, at the end of this phase, we had a working prototype concerning the main features. In the development phase we have been adopted the incremental and iterative software development methodology [18], consisting of new meetings with the future users where the features implemented can be discussed and refined and new features can be proposed. Since all the modules were implemented, we have began the deployment phase, which we divided into test stage and maintenance stage.
4. Results 3.1.4. Data integration module The Data Integration Module handles the batch insertion of data imported from other systems, manages coding standards and terminologies and performs the checking of data integrity. The batch insertion of patients data that come from other system can be done through a CSV (Comma-Separated Values) file. In the page for patient insertion, there is a button “Import Patients” that allows the user to specify the batch file. The button “CSV Guidelines” leads to details about the CSV file format. In the current version of the system, only patient's identification data can be imported. The data integrity checking is done mainly to avoid duplicity of information. When a new patient is registered in the system, the Data Integration Module checks key information (the combination of full name and birth date) to prevent duplication. The Data Integration Module uses the International Classification of Diseases revision 10 (ICD-10) [12] for inputs related to patient diagnosis. The ICD is an international standard endorsed by World Health Organization since 1994 and is used by a huge number of health related professionals such as physicians, nurses, insurers, and health information technology workers to classify diseases and other health problems. The system allows the user to enter the ICD code through the general information form. After ICD code validation, the primary diagnosis field is automatically filled with the description of the disease related to the ICD code, which ensures standardization and eliminates typos.
At the time of this publication, the clinical and biomaterial data storage and retrieval functions are fully implemented and can be accessed for demonstration at the website dcm.ffclrp.usp.br/caib/? pg¼biobanco. The system has been designed to serve different cancer treatment centers, which are responsible for a wide range of collection groups defined by tumor regions, such as central nervous system, lungs, urologic system, colorectal, etc. The main role of the collection groups is to add new data such as demographic information, clinical history, anatomopathological images and exams, and biomaterial from patients treated at cancer centers. This model enables the creation of research projects that can request biomaterial samples from the collection groups for use in cancer-related studies. The sociodemographic information about patients stored in the database includes: identification, address, occupation, birthplace, collection place, name of the medical group that collected the patient's clinical and biomaterial information, etc. The information on biomaterial includes: the patient from whom the biomaterial was extracted, processing status, type of sample (blood, plasma or tissue), type of post-processed sample (RNA, DNA or protein), quality of the biomaterial, place where it was collected, the medical group responsible for the sample, and whether it is a sub-sample of another sample. The clinical information comprises general information, such as patient family history, smoking and drinking history, as well as specific information about exams of the different types of tumor that are covered by the Clinical Genome project.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
In Fig. 3 we can see a list of patients, containing some basic individual information such as name, birth date and center where these data were collected. An important feature to be noticed in the interface is the presence of colored ellipsis at the right column. These ellipsis represent the patient's information that was already stored. Each one denotes a kind of information, such as identification, general information, images, etc. In addition, each ellipsis color has a different meaning: gray denotes the absence of that kind of information; yellow indicates incomplete information and green shows that information is complete. The interface also has a simple search field where the user can type patient's name to filter patient listing. It is important to note that the patients
5
listed in this interface are only the ones that the user has permission to access, granting the information confidentiality. Other action that can be done using the patients list is to add a new information set by clicking on the plus icon near the colored ellipsis. The system will ask the user what kind of patient information he/she wants to add and after the choice he/she will be led to that information form, as shown in Fig. 4 for clinical laboratory data. Finally, if the user needs to print any information set, there is a printing questionnaire button, where he/she will be able to select the form to be printed. As proposed, we implemented a biomaterial management environment, which enables the users of BBW to request and
Fig. 3. Interface showing the patient list with color badges representing the type of information available for those patients. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)
Fig. 4. Clinical laboratory data insertion interface.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
6
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Fig. 5. Biomaterial list.
Fig. 6. Biomaterial requisition interface.
provide biomaterial samples. Fig. 5 shows the main interface for this task. This interface provides the user with a biomaterial list which carries some information such as available volume, sample type, process status and if the biomaterial has sub aliquots, what is denoted by gray ellipsis. It also presents at the left hand a set of filters, which make it possible to refine the biomaterial listing according to user's needs. At the right we can see a group of icons that represent some actions that can be done by the user, such as: view, edit and delete biomaterial details, mark biomaterial as empty and request an aliquot. The aliquot requisition has its own interface, which can be seen in Fig. 6. The user can set the volume being requested, the project in which the aliquot will be used and some remarks. After the request, a coordinator will be asked to authorize the aliquot withdrawal, containing the released volume and other observations. The system automatically resets the volume of all aliquots
approved and not withdrawn after thirty days, allowing other users to request that biomaterial. In addition, all the biomaterial and its aliquots and sub aliquots can be displayed with more details as can be seen in Fig. 7. At left hand it is possible to see a biomaterial tree that shows to the user all sub samples extracted from the selected biomaterial, which are represented by the selected biomaterial's children.
5. Conclusions Our goal in this study was to deploy a system that enables members of cancer genome research groups to manage, store and merge their clinical, biomolecular and biomaterial data. This system allows for the storage and retrieval of clinical information about the patients whose biomaterial samples are stored in it. BBW possesses
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
7
Fig. 7. Biomaterial aliquots tree interface.
several functionalities to allow for management of these biomaterial samples. With regard to patient clinical data, BBW stores personal and sociodemographic information, as well as information on personal exams, images, follow-ups, etc., related to specific clinical fields of patient treatment and research. The system is already deployed to perform usability tests with a group of users of CISBi and Biobank Nucleus, providing them with data that will be analyzed to determine whether or not the system meets their requirements. A data analysis system is being implemented to allow users to analyze biomaterial data using specific samples and clinical data of patients from whom biomaterials were extracted. As future work, we plan to develop a new module called Semantic Annotation Module. This module will allow users to annotate the biomaterials and projects stored in the BBW, using concepts from biomedical ontologies, coding systems and terminologies. Currently, we have already implemented ICD-10 to annotate diagnosis. Other ontologies and terminologies that can be used in BBW are Gene Ontology (GO) to represent biological processes and gene functions studies, and Logical Observation Identifier Names and Codes (LOINC) to represent drugs in general. This semantic annotation is an important feature to improve the identification of relevant biomaterials that can be used in new research projects, leading to better standardization and interoperability. BBW enables research groups and projects to manage, make available, share and merge material and information in a practical and user-friendly way, improving the quality of specific genetic research. In this sense we intend to give a contribution to translational research.
Conflicts of interest statement None declared.
Acknowledgment The authors thank Brazilian Coordination for Improvement of Superior Level People (CAPES) for the financial support. References [1] A. Feldman, Does academic culture support translational research? Clin. Transl. Sci. 1 (2) (2008) 87–88. [2] S.H. Woolf, The meaning of translational research and why it matters, J. Am. Med. Assoc. 299 (2) (2008) 211–213. [3] A. Tan, B. Tripp, D. Daley, Brisk—research-oriented storage kit for biology related data, Bioinformatics (2011) 2422–2425. [4] T. Viangteeravat, et al., Biomedical informatics unit (BMIU): slim-prim system bridges the gap between laboratory discovery and practice, Clin. Transl. Sci. 2 (3) (2009) 238–241. [5] H.J. Lowe et al., STRIDE–an integrated standards-based translational research informatics platform, in: Proceedings of the AMIA Annual Symposium, the American Medical Informatics Association, 1, January 2009, pp. 391–395. [6] S.N. Murphy et al., Architecture of the open-source clinical research chart from informatics for integrating biology and the bedside, in: Proceedings of the AMIA Annual Symposium, the American Medical Informatics Association, 548–552, January 2007, pp. 548–552. [7] N.S.B. Miyoshi, et al., Computational framework to support integration of biomolecular and clinical data within a translational approach, BMC Bioinform. 14 (2013) 180. [8] P. Elliot, T.C. Peakman, The UK biobank sample handling and storage protocol for the collection, processing and archiving of human blood and urine, Int. J. Epidemiol. 37 (2007) 234–244. [9] K.S. Rønningen, L. Paltiel, H.M. Meltzer, R. Nordhagen, K.K. Lie, R. Hovengen, M. Haugen, W. Nystad, P. Magnus, J.A. Hoppin, The biobank of the Norwegian mother and child cohort Study: a resource for next 100 years, Eur. J. Epidemiol. 21 (2006) 619–625. [10] D.M. Roden, J.M. Pulley, M.A. Basford, G.R. Bernard, E.W. Clayton, J.R. Balser, D.R. Masys, Development of a large-scale de-identified DNA biobank to enable personalized medicine, Clin. Pharmacolo. Ther. 84 (2008) 362–369. [11] C.T. Fan†, J.C. Lin, C.H. Lee, Taiwan Biobank: a project aiming to aid Taiwan's transition into a biomedical island, Pharmacogenomics 9 (2) (2008) 235–246. [12] International Classification for Diseases, ICD Website, 〈http://www.who.int/ classifications/icd/en/〉 (last access on October 2014). [13] PostgreSQL, PostgreSQL Website, 〈http://www.postgresql.org/〉 (last access on January 2014). [14] Django, Django Website 〈https://www.djangoproject.com/〉 (last access on January 2014).
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
8
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
[15] Twitter Bootstrap, Twitter Bootstrap Website, 〈http://twitter.github.io/boot strap/〉 (last access on January 2014). [16] R.T. Fielding, G. Kaiser, The Apache HTTP server project, IEEE Internet Comput. 1 (4) (1997) 88–90. [17] J.M. Carey, Prototyping: alternative systems development methodology, Inf. Softw. Technol. 32 (2) (1990) 119–126. [18] C. Larman, V.R. Basili, Iterative and incremental developments. a brief history, Computer 36 (6) (2003) 47 56.
Yuri Ferretti received the B.Sc. degree in Biomedical Informatics from University of São Paulo, Brazil, in 2011. Presently he is a M.Sc. student in the Inter-Institutional Grad Program on Bioinformatics also from University of São Paulo. His research interests include bioinformatics, translational research and biological databases.
Wilson Araújo Silva Jr received the B.Sc. degree in Biology (Medical Modality) from Federal University of Pará, Brazil, in 1989, the M.Sc. degree in Genetics from University of Sao Paulo, Brazil, in 1993 and the Ph.D. degree also in Genetics from University of Sao Paulo, Brazil, in 1999. He is currently an Associated Professor in the Genetics Department of University of São Paulo at Ribeirão Preto, Brazil. He coordinates the Center of Genomic Medicine and the Integrated Systems Biology Research Nucleous of Medical School of University of São Paulo at Ribeirão Preto, Brazil. His research interests include bioinformatics, genomic medicine and cancer genetics.
Joaquim Cezar Felipe received the B.Sc. degree in Mechanical Engineering from University of São Paulo, Brazil, in 1986, the M.Sc. degree in Computer Science from Federal University of São Carlos, Brazil, in 2000 and the Ph.D. degree in Computer Science from University of Sao Paulo, Brazil, in 2005. He is currently a full Professor in the Computing and Mathematics Department of University of São Paulo at Ribeirão Preto, Brazil. His research interests include bioinformatics, biological databases and medical image processing and analysis.
Newton Shydeo Brandão Miyoshi received the B.Sc. degree in Biomedical Informatics from University of São Paulo, Brazil, in 2009 and the M.Sc. in Bioinformatics also from University of São Paulo, Brazil, in 2013. Presently he is a Ph.D. student in the Grad Program on Medical Clinic also from University of São Paulo. His research interests include bioinformatics, health informatics and data integration.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
Contents lists available at ScienceDirect
Computers in Biology and Medicine journal homepage: www.elsevier.com/locate/cbm
BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data Yuri Ferretti a,b, Newton Shydeo Brandão Miyoshi a, Wilson Araújo Silva Jrc,d, Joaquim Cezar Felipe a,b,c,n a
Inter-institutional Post-graduation Program on Bioinformatics, University of São Paulo, Brazil Department of Computing and Mathematics, Faculty of Philosophy Sciences and Languages of Ribeirão Preto, University of São Paulo, Brazil c Center for Integrative Systems Biology - CISBi, NAP/USP, University of São Paulo, Brazil d Department of Genetics, School of Medicine of Ribeirão Preto, University of São Paulo, Brazil b
art ic l e i nf o
a b s t r a c t
Article history: Received 30 April 2014 Accepted 4 April 2015
Background: Researchers of translational medicine face numerous challenges in attempting to bring research results to the bedside. This field of research covers a wide range of resources, including blood and tissue samples, which are processed for isolation of RNA and DNA to study cancer omics data (genomics, proteomics and metabolomics). Clinical information about patients' habits, family history, physical examinations, remissions, etc., is also important to underpin studies aimed at identifying patterns that lead to the development of cancer and to its successful treatment. Purpose: Development of a web-based computer system—BioBankWarden—to manage, consolidate and integrate these diversified data, enabling cancer research groups to retrieve and analyze clinical and biomolecular data within an integrative environment. The system has a three-tier architecture comprising database, logic and user-interface layers. Results: The system's integrated database and user-friendly interface allow for the control of patient records, biomaterial storage, research groups, research projects, users and biomaterial exchange. Conclusions: BioBankWarden can be used to store and retrieve specific information from different clinical fields linked to biomaterials collected from patients, providing the functionalities required to support translational research in the field of cancer. & 2015 Elsevier Ltd. All rights reserved.
Keywords: Translational medicine Biological databases Oncogenomics Biomaterial management Clinical data management
1. Introduction Human genome mapping and the emergence of new sequencing techniques have motivated cancer researchers to focus a great deal of attention on understanding the mechanisms that contribute to the development of oncology diseases. This includes translational medicine, which combines basic, social and political sciences to optimize patient care and preventive practices and to identify behavioral and environmental factors that can trigger mutations in the human genetic code, which can give rise to a variety of tumors [1]. Translational medicine involves the application of scientific research results, particularly those of omics techniques, to improve health and treatment processes [2], aiming to bridge the gap between research and clinical practice.
n Corresponding author at: Av Bandeirantes, 3900, Ribeirão Preto, SP, 14040-901, Brazil. E-mail addresses: [email protected] (Y. Ferretti), [email protected] (N.S.B. Miyoshi), [email protected] (W.A. Silva Jr), [email protected] (J.C. Felipe).
Scientists usually investigate abnormal levels of gene expression that could be related to specific cancer diseases in order to identify the gene responsible for the disorder. This requires the collection of tissue and blood samples from patients to isolate DNA, RNA and proteins, which can reveal genetic mutations. A concomitant analysis of patients' personal and clinical data, such as family history, smoking and drinking habits, race, medical history, etc., can also be performed to identify cancer-related factors. Finally, the data obtained from these different sources can be analyzed jointly to determine which environmental or behavioral factors play meaningful roles in cancer genesis. The purpose of this study was to develop a web-based computer system—called BioBankWarden (BBW)—which enables researchers to store, retrieve and integrate biomolecular and clinical data on biomaterials collected from patients, as recommended by Brisk research-oriented storage kit [3]. BBW supports the management of biomaterial data, such as requisitions and submissions of aliquots, to help researchers to determine the requirements for their studies. Another important point is to manage user permissions to access data from specific research groups, avoiding ethical issues. To facilitate the control and access of patient data from different clinical
http://dx.doi.org/10.1016/j.compbiomed.2015.04.008 0010-4825/& 2015 Elsevier Ltd. All rights reserved.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
2
areas, BBW provides different data sets for each clinical area, according to the user's needs. The Bank of Central Nervous System Tumors, which emerged as a subproject of the Clinical Genome Project, an initiative of the Department of Neurosurgery of Ribeirão Preto Medical School (RPMS) – University of São Paulo (USP) – Brazil, initially contained samples of only one type of nervous system tumor. This initial collection of biological material was later expanded to encompass all the tumors of the central nervous system, as well as other diseases treated by the neurosurgical team of the Teaching Hospital of RPMS, i.e., urology, gastric surgery, proctology and gynecology. The Biobank Nucleus for Research Support, which was created in 2011 through a collaboration of several researchers at the Translational Laboratory, also contains a large number of samples. In 2012, CISBi (Center for Integrative Systems Biology) was created to provide computational tools for the analysis, integration and interpretation of biological data and to encourage the formulation of new concepts involving relevant biological issues, such as dynamic systems theories, analysis of molecular noise, and statistics applied to genetic networks. The development of the BBW system is the result of the Biobank and CISBi merger. BBW will support the storage, management and integration of data from all the projects related to these initiatives. The researchers who participate in the projects contribute by detailing the system requirements and evaluating the prototypes. Several computational platforms have been designed to store and query biomolecular and clinical data. Some examples are: SlimPrim (Scientific Laboratory Information Management – Patient-care Research Information Management) [4]; STRIDE (Stanford Translational Research Integrated Database Environment) [5]; and I2B2 (Informatics for Integrating Biology and the Bedside) [6]. Miyoshi et al. [7] implemented the IPTrans (Integrative Platform for Translational Research), a framework for integrating biomolecular and clinical data based on the use of a common reference ontology. The clinical database of this platform is based on the Entity-AttributeValue data structure. However, this platform does not provide tools for controlling the storage and distribution of biomaterial samples or for discriminating specific clinical data according to different clinical areas (digestive tract, female reproductive tract, etc.) or for arranging data and users according to research groups and projects. IPTrans framework supports the integration of heterogeneous data sources containing personal, demographic and clinical information about patients. It has also a database to store results of biomolecular procedures, such as levels of genes expression, performed on the same patients' genetic material. In addition, an analytical module is under development, which will enable the user to perform analyses by merging clinical data and biomolecular results. As part of this environment, BBW will be a clinical data source of IPTrans. The results of biomolecular procedures performed on BBW biomaterials will be stored in the biomolecular database of IPTrans and then the analytical module will be able to perform the clinical/omics integrative analysis. Thus, the main benefit we expect from BBW is that it can act as a clinical data source for a broader environment, namely the IPTrans, thereby promoting the desired linking with biomolecular data, thus enabling the user to accomplish integrative analyses. Other major contribution of the system's implementation is its capability to tightly control storage and usage of biological materials. Furthermore, BBW enables the management of information about patients and biomaterials by assigning them to research projects and/or groups and also defining a user hierarchy of access in such a way to guarantee security and privacy of sensitive data.
2. Similar studies Elliot and Peakman [8] describe the UK Biobank system as a large database composed of biological samples collected from
500,000 patients between 2007 and 2010. The clinical data, which comprise physical and cognitive measurements, were collected by means of questionnaires, and the biological samples consist of urine and blood. All the samples are stored and managed by a proprietary LIMS (Laboratory Information Management System). In Norway, Ronning et al. [9] created a biobank with 380,000 biological samples comprising mainly whole blood, plasma, DNA and urine from pregnant women, their partners and their children. The LIMS manages de-identification, location of the aliquots in freezers, and quality control of the entire process. Roden et al. [10] developed a DNA biobank linked to clinical information from a de-identified electronic medical record (EMR) system. This biobank receives a constant supply of new DNA samples, at a rate of 700–900 samples per week. De-identification and sampling algorithms were developed and integrated into the biobank information system. In Taiwan [11], a biobank was created to collect DNA from a large group of people, aiming to track their health and lifestyle for at least 10 years. The participants will be volunteers recruited throughout the country, in the 30-year age range and of both sexes. The main goal is to track biological data on common chronic diseases such as high blood pressure, diabetes and cancer. Although other research groups have implemented biobank systems, most of these systems focus on specific diseases or conditions, e.g., the Norway Biobank, which focuses on obtaining biological material from pregnant women. Other systems focus on a specific type of biomaterial such as DNA. Another limitation of these biobanks is the fixed size of samples, such as the UK Biobank, which does not accept new biomaterials over time. In addition, patients' clinical and demographic information in these biobanks established a priori, and therefore does not allow for the addition of new clinical information. Unlike the above cited systems, the computational framework and functionalities of BBW, which are described below, appear to be more generic, robust and flexible than the existing solutions. Another relevant feature of BBW that differentiates it from other proposals is its capability to support the management of users, research groups and projects. In contrast to other systems, BBW allows for the creation and association of different projects involving different groups of users, and has a requisition module to manage the output of material based on roles and permissions for exchanges among collaborative groups.
3. Methodology To build BBW, we have been using standard methodologies for system design, development and deployment. Up-to-date technologies and open-source software enables data integration and access of users from different locations, using any web platform. 3.1. Architecture of BioBankWarden BBW is a web-based information system following a clientserver architecture. It is composed by a three-tier architecture: database, logic and user-interface tiers, respectively responsible for storing, processing and presenting the data (Fig. 1). The user-interface tier provides a friendly and intuitive interface that can be accessed from desktop computers, laptops and post-PC devices, such as tablets and smartphones. The database tier is responsible for storing and retrieving all data (clinical and biomaterial) related to the research projects. Finally, the business logic tier comprises four modules: Security module, Biomaterial Module, Data Integration Module and Clinical Module. In these modules all the methods and functions responsible for exchanging information between the database and the user-interface tiers are implemented.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
This tier also performs the validation of data inserted by the user to prevent inconsistencies and controls the user permissions for access, visualization and edition of data. 3.1.1. Security module The security module is responsible for implementing registration, authentication and authorization protocols. The research data stored in BBW demands the creation of many levels of access, restricting or enabling access to the system functionalities. One important assignment of BBW is to allow independent research groups to search biomaterials that could be relevant to their ongoing projects. Any researcher can register to access the system and search for biomaterials of interest and then request these biomaterials. On the other side, hospitals and biomedical centers can also dispose their biomaterials to be managed and accessed through BBW. To provide this flexibility and also guarantee privacy of patient sensitive data we defined a hierarchy of access. This hierarchy allows users to be classified into two distinct roles in the system: users belonging to a disease research group and independent researchers. The users belonging to a disease research group can register new biomaterials and are responsible for the clinical information UI
Independent Researcher
Member Researcher
User interface tier
Security Module Business logic tier Biomaterial Module
Data Integration Module
Clinical Module
Database tier
Database BioBankWarden Fig. 1. Conceptual schema of BBW architecture.
3
about the patients associated to the group. Within a research group there are three types of users: coordinator, medical resident and member. The coordinator can invite new users to take part in the research group, managing group members. The medical resident is responsible for registering patient clinical information and biological samples associated with the patient. Members only have the permission to visualize the clinical data. The other main user, the independent researcher, can ask for the creation of projects and consequently seek and request biomaterials. Within a project there are the following roles: project coordinator, who performs the management of the members of this project; resident, who can make requests of biological samples; and guest, who can view the documents and information from biological samples associated with the projects, but cannot request or edit any information associated to biomaterials. This division into two main groups is also important to guarantee patient privacy. Independent research users are not allowed to access the patient sensitive information, but only those associated to the biomaterial. In addition to these users types there are the curators, responsible for assigning samples to projects and controlling them, and the system administrators, who have full control of the system and can register new research groups and manage clinical surveys. This user hierarchy schema is shown in Fig. 2.
3.1.2. Clinical module BBW is flexible on dealing with various types of biological samples originated from different cancer studies. Each group of patients can have different sets of clinical information available. This information is stored in BBW through Clinical Surveys managed by the Clinical Module. Each disease research group is responsible for providing the clinical information about their patients. The resident physicians that belong to some disease research group is the user responsible for doing the patient clinical data registration using a set of clinical surveys. The surveys are divided into two types: general and research. The general surveys are required for all disease research groups. There are two general surveys representing identification and clinical
Fig. 2. BBW user hierarchy schema.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
4
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
information about the patients: identification survey contains information such as documents, address and contact; and clinical survey contains information such as age, weight, height, family history, use of alcohol, smoking history, etc. The research survey forms vary according to the domain of interest of the disease research groups. Each group can share or create surveys themselves. Currently there are four research surveys: clinical laboratory, anatomic pathology, imaging examination and follow-up. All of them are related to the oncology domain. The clinical laboratory survey represents general information about the tumor, the preoperative, surgery and postoperative stages. The anatomic pathology survey represents more specific information on tumor type, diagnosis and gene expression. The follow-up survey aggregates information about the treatment applied (chemotherapy, radiotherapy) and its output (cure, death, recurrence). The image examination survey allows the addition of images associated to the tumor.
3.2. Technology BBW is built by Web 2.0 technologies, allowing fast prototyping, flexible interface and robust data management. It uses PostgreSQL (version 9.0) [13] as the relational database management system. The logic and user-interface tiers were built using the python object oriented web development framework called Django (version 1.4) [14]. BBW follows the Model-View-Controller (MVC) pattern for software design, enabling clear separation from business logic through declaration of Models, user-interface with the Views and the information flow management in Controllers. To build a clean and ease-to-use interface we have used the library Twitter-bootstrap (version 2.3.2) [15], a front-end web development framework that helps developers to easily and quickly design web pages. The system is deployed with the Apache HTTP Server [16] in Ubuntu Server version 14.04. 3.3. Development strategy
3.1.3. Biomaterial module The Biomaterial Module is responsible for the most important assignment of BBW: the biomaterial management. It allows the system user to control storage, retrieval and access to biomaterial samples. The system accepts several types of samples like blood, serum, tumor margin and others. The volume of a stored sample can be fractioned in smaller amounts, in order to facilitate the storage or distribution among researchers. In the Biomaterial Module the user can also generate processed samples, i.e. DNA, RNA and protein, extracted from the original biomaterial samples. Another requirement is to control how biomaterial samples are sent to researchers that request it. The system allows researchers to request amounts of biomaterial samples, but a coordinator from the owner disease research group must approve each request. It is also possible to filter results when the user searches for stored biomaterials that were extracted from patients with specific characteristics, such as age, weight, smoking history and gender, among others.
The software development life cycle of BBW is divided into three phases: design phase, development phase and deployment phase. In each phase we have adopted a different methodology. In the design phase we have followed the prototyping software development methodology [17]: initial meetings with research groups to analyze the system requirements such as the clinical data to be stored, the users roles for viewing and modifying data and the biomaterial relevant data. Thus, at the end of this phase, we had a working prototype concerning the main features. In the development phase we have been adopted the incremental and iterative software development methodology [18], consisting of new meetings with the future users where the features implemented can be discussed and refined and new features can be proposed. Since all the modules were implemented, we have began the deployment phase, which we divided into test stage and maintenance stage.
4. Results 3.1.4. Data integration module The Data Integration Module handles the batch insertion of data imported from other systems, manages coding standards and terminologies and performs the checking of data integrity. The batch insertion of patients data that come from other system can be done through a CSV (Comma-Separated Values) file. In the page for patient insertion, there is a button “Import Patients” that allows the user to specify the batch file. The button “CSV Guidelines” leads to details about the CSV file format. In the current version of the system, only patient's identification data can be imported. The data integrity checking is done mainly to avoid duplicity of information. When a new patient is registered in the system, the Data Integration Module checks key information (the combination of full name and birth date) to prevent duplication. The Data Integration Module uses the International Classification of Diseases revision 10 (ICD-10) [12] for inputs related to patient diagnosis. The ICD is an international standard endorsed by World Health Organization since 1994 and is used by a huge number of health related professionals such as physicians, nurses, insurers, and health information technology workers to classify diseases and other health problems. The system allows the user to enter the ICD code through the general information form. After ICD code validation, the primary diagnosis field is automatically filled with the description of the disease related to the ICD code, which ensures standardization and eliminates typos.
At the time of this publication, the clinical and biomaterial data storage and retrieval functions are fully implemented and can be accessed for demonstration at the website dcm.ffclrp.usp.br/caib/? pg¼biobanco. The system has been designed to serve different cancer treatment centers, which are responsible for a wide range of collection groups defined by tumor regions, such as central nervous system, lungs, urologic system, colorectal, etc. The main role of the collection groups is to add new data such as demographic information, clinical history, anatomopathological images and exams, and biomaterial from patients treated at cancer centers. This model enables the creation of research projects that can request biomaterial samples from the collection groups for use in cancer-related studies. The sociodemographic information about patients stored in the database includes: identification, address, occupation, birthplace, collection place, name of the medical group that collected the patient's clinical and biomaterial information, etc. The information on biomaterial includes: the patient from whom the biomaterial was extracted, processing status, type of sample (blood, plasma or tissue), type of post-processed sample (RNA, DNA or protein), quality of the biomaterial, place where it was collected, the medical group responsible for the sample, and whether it is a sub-sample of another sample. The clinical information comprises general information, such as patient family history, smoking and drinking history, as well as specific information about exams of the different types of tumor that are covered by the Clinical Genome project.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
In Fig. 3 we can see a list of patients, containing some basic individual information such as name, birth date and center where these data were collected. An important feature to be noticed in the interface is the presence of colored ellipsis at the right column. These ellipsis represent the patient's information that was already stored. Each one denotes a kind of information, such as identification, general information, images, etc. In addition, each ellipsis color has a different meaning: gray denotes the absence of that kind of information; yellow indicates incomplete information and green shows that information is complete. The interface also has a simple search field where the user can type patient's name to filter patient listing. It is important to note that the patients
5
listed in this interface are only the ones that the user has permission to access, granting the information confidentiality. Other action that can be done using the patients list is to add a new information set by clicking on the plus icon near the colored ellipsis. The system will ask the user what kind of patient information he/she wants to add and after the choice he/she will be led to that information form, as shown in Fig. 4 for clinical laboratory data. Finally, if the user needs to print any information set, there is a printing questionnaire button, where he/she will be able to select the form to be printed. As proposed, we implemented a biomaterial management environment, which enables the users of BBW to request and
Fig. 3. Interface showing the patient list with color badges representing the type of information available for those patients. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)
Fig. 4. Clinical laboratory data insertion interface.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
6
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Fig. 5. Biomaterial list.
Fig. 6. Biomaterial requisition interface.
provide biomaterial samples. Fig. 5 shows the main interface for this task. This interface provides the user with a biomaterial list which carries some information such as available volume, sample type, process status and if the biomaterial has sub aliquots, what is denoted by gray ellipsis. It also presents at the left hand a set of filters, which make it possible to refine the biomaterial listing according to user's needs. At the right we can see a group of icons that represent some actions that can be done by the user, such as: view, edit and delete biomaterial details, mark biomaterial as empty and request an aliquot. The aliquot requisition has its own interface, which can be seen in Fig. 6. The user can set the volume being requested, the project in which the aliquot will be used and some remarks. After the request, a coordinator will be asked to authorize the aliquot withdrawal, containing the released volume and other observations. The system automatically resets the volume of all aliquots
approved and not withdrawn after thirty days, allowing other users to request that biomaterial. In addition, all the biomaterial and its aliquots and sub aliquots can be displayed with more details as can be seen in Fig. 7. At left hand it is possible to see a biomaterial tree that shows to the user all sub samples extracted from the selected biomaterial, which are represented by the selected biomaterial's children.
5. Conclusions Our goal in this study was to deploy a system that enables members of cancer genome research groups to manage, store and merge their clinical, biomolecular and biomaterial data. This system allows for the storage and retrieval of clinical information about the patients whose biomaterial samples are stored in it. BBW possesses
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
7
Fig. 7. Biomaterial aliquots tree interface.
several functionalities to allow for management of these biomaterial samples. With regard to patient clinical data, BBW stores personal and sociodemographic information, as well as information on personal exams, images, follow-ups, etc., related to specific clinical fields of patient treatment and research. The system is already deployed to perform usability tests with a group of users of CISBi and Biobank Nucleus, providing them with data that will be analyzed to determine whether or not the system meets their requirements. A data analysis system is being implemented to allow users to analyze biomaterial data using specific samples and clinical data of patients from whom biomaterials were extracted. As future work, we plan to develop a new module called Semantic Annotation Module. This module will allow users to annotate the biomaterials and projects stored in the BBW, using concepts from biomedical ontologies, coding systems and terminologies. Currently, we have already implemented ICD-10 to annotate diagnosis. Other ontologies and terminologies that can be used in BBW are Gene Ontology (GO) to represent biological processes and gene functions studies, and Logical Observation Identifier Names and Codes (LOINC) to represent drugs in general. This semantic annotation is an important feature to improve the identification of relevant biomaterials that can be used in new research projects, leading to better standardization and interoperability. BBW enables research groups and projects to manage, make available, share and merge material and information in a practical and user-friendly way, improving the quality of specific genetic research. In this sense we intend to give a contribution to translational research.
Conflicts of interest statement None declared.
Acknowledgment The authors thank Brazilian Coordination for Improvement of Superior Level People (CAPES) for the financial support. References [1] A. Feldman, Does academic culture support translational research? Clin. Transl. Sci. 1 (2) (2008) 87–88. [2] S.H. Woolf, The meaning of translational research and why it matters, J. Am. Med. Assoc. 299 (2) (2008) 211–213. [3] A. Tan, B. Tripp, D. Daley, Brisk—research-oriented storage kit for biology related data, Bioinformatics (2011) 2422–2425. [4] T. Viangteeravat, et al., Biomedical informatics unit (BMIU): slim-prim system bridges the gap between laboratory discovery and practice, Clin. Transl. Sci. 2 (3) (2009) 238–241. [5] H.J. Lowe et al., STRIDE–an integrated standards-based translational research informatics platform, in: Proceedings of the AMIA Annual Symposium, the American Medical Informatics Association, 1, January 2009, pp. 391–395. [6] S.N. Murphy et al., Architecture of the open-source clinical research chart from informatics for integrating biology and the bedside, in: Proceedings of the AMIA Annual Symposium, the American Medical Informatics Association, 548–552, January 2007, pp. 548–552. [7] N.S.B. Miyoshi, et al., Computational framework to support integration of biomolecular and clinical data within a translational approach, BMC Bioinform. 14 (2013) 180. [8] P. Elliot, T.C. Peakman, The UK biobank sample handling and storage protocol for the collection, processing and archiving of human blood and urine, Int. J. Epidemiol. 37 (2007) 234–244. [9] K.S. Rønningen, L. Paltiel, H.M. Meltzer, R. Nordhagen, K.K. Lie, R. Hovengen, M. Haugen, W. Nystad, P. Magnus, J.A. Hoppin, The biobank of the Norwegian mother and child cohort Study: a resource for next 100 years, Eur. J. Epidemiol. 21 (2006) 619–625. [10] D.M. Roden, J.M. Pulley, M.A. Basford, G.R. Bernard, E.W. Clayton, J.R. Balser, D.R. Masys, Development of a large-scale de-identified DNA biobank to enable personalized medicine, Clin. Pharmacolo. Ther. 84 (2008) 362–369. [11] C.T. Fan†, J.C. Lin, C.H. Lee, Taiwan Biobank: a project aiming to aid Taiwan's transition into a biomedical island, Pharmacogenomics 9 (2) (2008) 235–246. [12] International Classification for Diseases, ICD Website, 〈http://www.who.int/ classifications/icd/en/〉 (last access on October 2014). [13] PostgreSQL, PostgreSQL Website, 〈http://www.postgresql.org/〉 (last access on January 2014). [14] Django, Django Website 〈https://www.djangoproject.com/〉 (last access on January 2014).
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
8
Y. Ferretti et al. / Computers in Biology and Medicine ∎ (∎∎∎∎) ∎∎∎–∎∎∎
[15] Twitter Bootstrap, Twitter Bootstrap Website, 〈http://twitter.github.io/boot strap/〉 (last access on January 2014). [16] R.T. Fielding, G. Kaiser, The Apache HTTP server project, IEEE Internet Comput. 1 (4) (1997) 88–90. [17] J.M. Carey, Prototyping: alternative systems development methodology, Inf. Softw. Technol. 32 (2) (1990) 119–126. [18] C. Larman, V.R. Basili, Iterative and incremental developments. a brief history, Computer 36 (6) (2003) 47 56.
Yuri Ferretti received the B.Sc. degree in Biomedical Informatics from University of São Paulo, Brazil, in 2011. Presently he is a M.Sc. student in the Inter-Institutional Grad Program on Bioinformatics also from University of São Paulo. His research interests include bioinformatics, translational research and biological databases.
Wilson Araújo Silva Jr received the B.Sc. degree in Biology (Medical Modality) from Federal University of Pará, Brazil, in 1989, the M.Sc. degree in Genetics from University of Sao Paulo, Brazil, in 1993 and the Ph.D. degree also in Genetics from University of Sao Paulo, Brazil, in 1999. He is currently an Associated Professor in the Genetics Department of University of São Paulo at Ribeirão Preto, Brazil. He coordinates the Center of Genomic Medicine and the Integrated Systems Biology Research Nucleous of Medical School of University of São Paulo at Ribeirão Preto, Brazil. His research interests include bioinformatics, genomic medicine and cancer genetics.
Joaquim Cezar Felipe received the B.Sc. degree in Mechanical Engineering from University of São Paulo, Brazil, in 1986, the M.Sc. degree in Computer Science from Federal University of São Carlos, Brazil, in 2000 and the Ph.D. degree in Computer Science from University of Sao Paulo, Brazil, in 2005. He is currently a full Professor in the Computing and Mathematics Department of University of São Paulo at Ribeirão Preto, Brazil. His research interests include bioinformatics, biological databases and medical image processing and analysis.
Newton Shydeo Brandão Miyoshi received the B.Sc. degree in Biomedical Informatics from University of São Paulo, Brazil, in 2009 and the M.Sc. in Bioinformatics also from University of São Paulo, Brazil, in 2013. Presently he is a Ph.D. student in the Grad Program on Medical Clinic also from University of São Paulo. His research interests include bioinformatics, health informatics and data integration.
Please cite this article as: Y. Ferretti, et al., BioBankWarden: A web-based system to support translational cancer research by managing clinical and biomaterial data, Comput. Biol. Med. (2015), http://dx.doi.org/10.1016/j.compbiomed.2015.04.008i
