Biomedical Informatics - Services
Biomedical Informatics Support for Genome Scale Analyses: Next Generation Sequencing (NGS) promises to revolutionize biomedical research. Through technological advances based on massive parallelization, NGS provides an enormous number of reads and permits sequencing of entire genomes (and their transcriptome). The Biomedical Informatics Shared Resource made considerable investment in computational hardware and Biomedical Informatics software development to manage the very large data sets generated and to extract biomedical insights from the data collected. We have pipelines to analyze data collected in RNA-Seq, Chip-Seq, whole genome sequencing, etc. using in-house shared memory computational platforms as well as deployed massively parallel sequence alignment/assembly codes on our University’s Newton MPP supercomputer. Biomedical Informatics deployed, and is using, sequence assembly, alignment codes (SOAP, Bowtie, Abyss) and data visualization tools (Galaxy, IGV from the Broad Institute) on shared memory and massively parallel supercomputers to analyze and visualize the vast data set generation from the Illumina and ABI/Solid Next Generation sequencers.
Microarray Database and Analysis: The Biomedical Informatics Shared Resource supports experimental design and analysis of microarray data, often generated by the Functional Genomics Shared Resource, including pathway analysis and molecular modeling. We expect this effort to lead to co-authorships in publications in the near future and requests for Biomedical Informatics support in future grant applications. In addition, we have deployed and manage data backup/storage (via a dedicated server and disk/tape devices) to ensure that the data collected in a PI laboratory is automatically copied to dedicated server storage at the Rutgers Cancer Institute's Data Center every night. The data is migrated monthly to an encrypted tape facility and shipped to a remote site.
CINJ Warehouse Services/Integrative Cancer Biology and Data Mining: Biomedical research has yet to fully harness the transformative power of information technology to enhance research productivity and efficiency, which accelerates research discoveries that transform clinical practice. Effective aggregation and management of knowledge along with data resources is critical to advancing clinical and translational science. It is a priority of the Biomedical Informatics Shared Resource to organize and implement informatics initiatives and their associated cyber-infrastructure and support capabilities to meet these needs. As in most medical centers, the information management needs of the combined clinical and translational research community at Rutgers Cancer Institute have historically been met by resources available within individual research programs that were too diffuse and/or lacked focus. In the past two years, Rutgers Cancer Institute leadership and individual investigators have been strong advocates and supporters of the Biomedical Informatics Shared Resource in its efforts to develop and deploy a model of information integration and data sharing to catalyze the translation of research discoveries and to advance research into quantifiable outcomes across traditional institutional and geographic boundaries. The Biomedical Informatics Shared Resource is working with Rutgers Cancer Institute investigators to facilitate mining of existing data based on each research project’s scientific priorities. Therefore, a significant component of the resource includes detailed outreach activities that are designed to maximize the power of the Biomedical Informatics Shared Resource’s expertise and resources to advance research. A major component of the efforts in support of integrative cancer biology is focused on developing, deploying and supporting data repositories using commercial and/or open source software that meet the specific requirements of Rutgers Cancer Institute members. The Biomedical informatics team has purchased and is rolling out an enterprise-wide clinical data warehouse based the web-accessible clinical & translational research management software system, LabMatrix offered from BioFortis, Columbia, MD. The data repositories will provide federated access to clinical data, archive research datasets from completed studies, link research data sources for multidisciplinary collaboration, and serve as a platform for translational research. Targeted clinical data sources for phase I of the project include the ARIA-EMR containing encounter, laboratory, other EMR data, computerized physician order entry and data originating from radiology reports, pathology reports, surgical notes, clinical history, nursing notes, genomic sequencing studies, and information resident in Oncore which supports ongoing clinical trials. Examples of research data sources that will be integrated into the data warehouse in phase II of the project include radiology PACS, digitized pathology repositories, and epidemiologic data from Rutgers Cancer Institute Shared Resources, NJ State Cancer Registry/SEER and data bases linking Rutgers Cancer Institute’s Network of Hospitals, etc. As part of data repository development, wherever possible, the Biomedical Informatics Shared Resource is implementing a standard format to which data from heterogeneous sources will be transformed for further use. Standardization will facilitate data sharing both within and outside Rutgers Cancer Institute. The data structures within repositories are searchable and graphically browsable with data elements, volume and date range of the source, as well as standard and lexical reference to the data element (such as DICOM, HL7 v2.x, SNOMED, UMLS, caDSR, ICD10) where applicable.
Clinical Data Repository and Services: There is a need for a data repository for the population sciences, epidemiologists and clinicians in the various disease specific groups at Rutgers Cancer Institute to view populations and disease trends of patients. To support phase I studies and NCI-investigator driven clinical trials at Rutgers Cancer Institute, the Biomedical Informatics Shared Resource works with Rutgers Cancer Institute’s Office of Human Research Services shared resource in providing hardware, software and disaster recovery support for their OnCore Clinical Trial Data Management System. They are working with the commercial vendor to expand on the capabilities for real-time electronic data capture from our Aria-EMR as well as expanding on the OnCore Biospecimen Module that will facilitate integration with the data capture efforts in our Biospecimen Repository Service Shared Resource. The Biomedical Informatics Shared Resource ensures that the right data in the right format flows into appropriate data repositories in an efficient and secure manner. They have supported the Rutgers Cancer Institute membership using CAISIS (see http://www.caisis.org), which is an open source, web-based, cancer data management system that integrates research with patient care. CAISIS was developed by Memorial Sloan-Kettering Cancer Center and is now utilized by many of the Comprehensive Cancer Centers for their own research as well as for the exchange of data to create larger populations and cohorts for study.
Histopathology Imaging and Computational Image Analysis: Digital pathology technology enables biomedical researchers to digitize de-identified patient specimens and experimental specimens for easy storage, duplication, sharing and analysis. With the newly acquired high-throughput Olympus VS120 whole slide scanner, the Imaging Shared Resource can help users capture their experimental results in high-quality images for publication and analysis. The four-color fluorescent imaging capability allows fluorescent immunohistochemistry specimens to be digitized in their entirety, avoiding common fluorescent study pitfalls such as sampling error and photo-bleaching. State-of-art multispectral imaging not only captures entire emission spectrum of specimens in small wavelength intervals but also supports further fluorescent multiplexing by providing a means to separate emission signals originating from each fluorophors. The Imaging Shared Resource also provides consultation and help to projects related to imaging and image analysis to users. Researchers with specific analytical needs can receive custom analysis and software development service by working with staff at the service.
Chemical Informatics Analysis: Informaticians use small molecule/peptide databases for 'in silico' screening studies of key enzymes and receptors to identify lead targets for design and development of novel therapeutics. Molecular dynamics techniques are applied to develop a broader understanding of the biophysical significance of mutation and in the estimation of drug-receptor complex binding free energies. Software packages used for docking (in silico screening) include Autodock, UCSF DOCK, Gold and Vina. Software used for molecular dynamics projects includes Amber, Gromacs and NAMD. Quantum mechanics programs like Gaussian and Spartan are used to study physical properties of small molecule drugs, small peptides and unusual nucleic acids. The resource uses the Modeller program for homology modeling and the Rosetta package to model small proteins via ab initio protein folding techniques. This service is a key component of the translational science efforts at Rutgers Cancer Institute.
Support of Other Rutgers Cancer Institute of New Jersey Shared Resources: The Biomedical Informatics Shared Resource develops the web portals for access to the services of most Rutgers Cancer Institute of New Jersey Shared Resources. In addition, more specific Biomedical Informatics needs of various Shared Resources are met, as follows:
- Biometrics Shared Resource (Weichung Joe Shih, Ph.D., Director) - The Biomedical Informatics Shared Resource provides (1) software engineering support and web portal development and maintenance; for example, the Biometrics Shared Resource required conversion of a complex Fortran program based on a Biomedical Informatics statistical algorithm to Java to ensure greater scalability and portability; (2) system administration support to maintain system and application software and trouble shoot the desktop workstations used for statistical analysis by members of the Biometrics Shared Resource; and (3) high performance computing resources and scientific computing expertise on multi-core massively parallel supercomputers.
- Computational Imaging (Dimitris Metaxas, Ph.D. Director and John Nosher, Clinical Co-Director) - The newly established computational imaging division at the Biomedical Informatics Shared Resource provides software and algorithm development to support quantitative analysis of imaged pathology and radiology for both clinical and pre-clinical imaging applications. These resources and tools will enable Rutgers Cancer Institute investigators to detect, quantify and track tumor response to therapy in clinical trial settings, with the overall goal of stratifying patient populations, guiding clinical decisions and improving therapy planning.
- Clinical and Research IT (John Kerrigan, Ph.D., Director) - Planning and implementation of a robust clinical data warehouse and data mining and discovery tools to support Precision Medicine applications including development of interfaces to pathology reports, electronic patient records, collaborators and vendors. Maintaining collaborator and vendor relationships. Developing processes and workflows.
- Functional Genomics Shared Resource (Shridar Ganesan, MD, PhD, Director) - We are in the midst of a major expansion of the Functional Genomics Shared Resource to accommodate the rapidly growing needs of our membership for Next-Generation Sequencing technologies. The Biomedical Informatics Shared Resource will deploy scalable, efficient and secure storage and archiving capabilities to ensure seamless flow from data acquisition to storage and processing. This involves increasing the existing computing resources within the Biomedical Informatics Shared Resource to meet immediate needs (currently 30TB and projected to grow to 100TB within a year). Based on the anticipated growth in storage needs and the costs involved in deploying and maintaining an in-house solution, we are also exploring new avenues (i.e. based on Cloud Computing) as a cost-efficient, financially and operationally viable option. The Biomedical Informatics Shared Resource also provides data analysis for users of the Functional Genomics Resource.
- Biospecimen Repository Service Shared Resource (Lorna Rodriguez, MD, PhD, Director) - The availability of annotated specimens is enhanced by leveraging data available from our new EMR as well as from the NJ State Tumor Registry (NJSTR).
- Histopathology Shared Resource (Lauri Goodell, M.D., Director and David Foran, Ph.D., Co-Director) - The Biomedical Informatics Shared Resource provides cyber-infrastructure advice and support for the Telepathology Project, linking instrumentation for remote viewing of organs for transplant within a facility located at the Robert Wood Johnson University Hospital. In addition, informatics and software engineering experts at Dr. Foran’s Bioimaging Center are using the Biomedical Informatics Shared Resource HPC resources located at the Rutgers Cancer Institute's data center for code development and large-scale simulations. The Biomedical Informatics Shared Resource is working in partnership with Dr. Foran’s team to develop a large-scale storage facility to support research needs in pathology informatics in general and the data generated by the tissue microarray service. The Biomedical Informatics Shared Resource will maintain the hardware, provide system administration support and will be responsible for data backup and disaster recovery of all pathology imaging data.
- Office of Human Research Services (Tracie K. Saunders, RN, MS, Director) - The Biomedical Informatics Shared Resource provides cyber-infrastructure advice and support for the Oncore Clinical Trials Software, a caBIG compatible commercial CTMS system that supports the data collection/management and study protocols for Rutgers Cancer Institute’s clinical trials. In addition, Oncore includes a software module that supports the data management for Rutgers Cancer Institute’s Biospecimen Repository Service Shared Resource. Our Database Architect worked with the company (Forte Research Systems) and biorepository colleagues to migrate the data stored in a homemade Perl-based software into the Oracle-based OnCore module. Our HPC supervisor worked with the company to migrate the entire OnCore application and legacy data to new hardware, in the process implementing a failover mechanism by reconfiguring the hardware so that two of the servers are located in our data center and the other two identical servers at the Robert Wood Johnson University Hospital Data Center (with appropriate tape backup for disaster recovery). This would ensure 24/7 availability of data to OHRS and minimize loss of data in case of hardware or network failure.
- Epidemiology Services Shared Resource (Antoinette M. Stroup, PhD, Director) - The Biomedical Informatics Shared Resource provides the newly developing Epidemiology Services Shared Resource with infrastructure and design support for research tracking and analytical databases. Specifically, in support of all ESC-run population-based investigations of cancer prevention, etiology, treatment and outcomes, the Biomedical Informatics Shared Resource houses and maintains an application server (SunFire x4140) equipped with 2X6-core Opteron cpus, 32GB of RAM and 4x146GB disks linked to a back-up application server; and a storage server with 18 terabytes of secure, encrypted storage linked to a duplicate back-up server. The application server can be used for statistical analysis of data sets using R or SAS. The storage server provides up to 18 terabytes of secure, encrypted storage as noted. In addition, the developing Epidemiology Services Shared Resource has access to the Biomedical Informatics Shared Resource secure, encrypted web server for hosting various research projects/initiatives. As new study-specific needs arise for data base design and access controls, the Biomedical Informatics Shared Resource works with Epidemiology Support Services staff to meet these needs and support ongoing work.
Web-based Application Development: It is difficult for many projects to separate web development (web site) from medical/clinical informatics and database design/installation since all these applications are through a web portal requiring a “website/interface” with underlying database design, programming and linkage to informatics. While a majority of the web portals are used to support ongoing outreach and educational initiatives, a significant number are targeted to the collection and integration of data in support of research in population science and translational research where integration of data from the lab, clinic and population based studies will advance our research agenda in cancer prevention, control and survivorship. At a practical level, the resources and manpower devoted to developing and deploying web portals benefits our research agenda at Rutgers Cancer Institute as it helps to increase recruitment for clinical trials carried out at the center by making patients and clinicians aware of the ongoing trials. Examples of projects implemented and maintained by the Bioinformatics Shared Resource include the following:
- Network Hospital Portal - The Rutgers Cancer Institute of New Jersey Network includes 16 hospitals across the State that provides cancer care to over one third of New Jerseyans. Each Network hospital offers their patients access to the latest cancer therapies and state-of-the-art cancer care available only at NCI-designated Cancer Centers and their Networks. This database connects the affiliated hospitals via web: http://cinjweb.cinj.rutgers.edu/network/content/about/
- Investigator Toolkit - Easy access to various tools, documents, and processes for Clinical Trial submission. This site can be accessed internally at: http://intranet.cinj.org/OHRS/Pages/Clinical-Investigator-Toolkit.aspx
- Honest Broker – Resource for Honest Broker services and application forms that can only be accessed internally. See http://intranet.cinj.org/Research/Pages/Honest-Broker-Program.aspx