Databases data integrity

Due to the critical nature of the data stored in the LIMS database, data integrity must be assured. Current technology allows for disk mirroring to ensure that any database additions and changes (and therefore changes to electronic records) are copied to additional, and possibly remote, locations. The function and use of each mirror for backup, for disaster recovery, and for system startup after shutdown must be understood. Each of these scenarios must be tested as part of the validation. 

The data model is highly normalised, which suits data integrity and manipulation via the SQL language, but creates difficulties for the end-user in terms of accessing the data. Straightforward queries may involve numerous tables which have to be joined together. As a practical work-around, the database includes several very large denormalized tables. 

Letunic, 1., et al., SMART 4.0 towards genomic data integration. Nucleic Acids Res, 2004, 32 Database issue, D142-4. 

Before reaching the point of complete data integration as given above, there are intermediary levels of data integration that are beneficial to better analysis of data from process analyzers. The best case would be to have all the data in a human readable form that is independent of the application data format. Over the years several attempts have been made to have a universal format for spectroscopic data, including JCAMP-DX and extensible markup language (XML). Because many instrument vendors use proprietary databases, and there is not a universal standard, the problem of multiple data formats persists. This has led to an entire business of data integration by third parties who aid in the transfer of data from one source to another, such as between instruments and the plant s distributed control system (DCS). 

Bioinformatics will play a vital role in overcoming this data integration and modeling challenge, because databases, visualization software, and analysis software must be built to enable data assimilation and to make the results accessible and useful for answering biological questions (p. 254). 

Influenza virus Influenza Research University of Texas Southwestern Medical Center Database Comprehensive, integrated data about influenza virus genome sequences, virus phenotypic characteristics, and results from surveillance activities for the discovery and development of influenza virus vaccines, diagnostics, and therapeutics (http //www.fludb.org/brc/home. do decorator=influenza)... 

So, what level of data integrity is acceptable The total percentage of errors we found in our study was 5%. Is this satisfactory Is this the best we can hope for Hopefully not, especially as more people become dependent on databases and the rate of production of data becomes ever faster. Clearly, there is a need for a system that will better validate data being entered in the most used CAS databases. It is desirable that the quality of the databases increases at the same time as they are mushrooming in size. 

Security protection is provided to assure data integrity. Database access is controlled to prevent unauthorized user access (for example, to sensitive areas) and to prevent unauthorized operations (for example, delete a record). 

Electronic records include data-storing memory cards used in embedded control systems such as steam sterilizers. These cards are used to transfer data (critical parameters, cycle times, etc.) from the control system to a main database application which is usually networked. A memory card is about the size of a credit card. Manual installation, removal, and transfer of the memory cards must be conducted in accordance with defined procedures. Reuse of cards must be carefully considered data must not be deleted until confirmation is received that it has been successfully downloaded to the database. The local operating enviromnent and the possibility of EMl/ESD damage to data integrity of the memory card must also be evaluated. 

Given the orientation of Part 11 toward data integrity, the FDA has stressed in many recent Warning Letters that it considers backup to be an integral part of protecting data integrity. For databases,... 

Warning Letters issued by the FDA since 1997 can be found on the FDA Web site at http // www.fda.gov/ foi/warning.htm. While there are many examples related to backup and most do not mention databases specifically, the following examples all serve to illustrate agency concern with the concept of backup as a critical part of guaranteeing data integrity ... 

Such a situation offers both benefits and challenges. Parallel testing, i.e., the running of two systems in parallel and comparing the outputs, offers the opportunity of increasing the level of confidence that the system meets all expectations and user requirements. However, at some point, cut-over to sole operation of the new or modified database will have to be achieved. A crucial consideration here is the assurance that all records from the previous system (be they manual, from a different database, or from a smaller version of the new database) are available for further processing, inspection, and manipulation, and most importantly, have had their data integrity preserved. This is usually achieved by validation of the process used to transfer the data into the new system. 

Cleaning and Transforming Data. When importing data from diverse data sources (files, databases, spreadsheets, LIMS systems, etc.) into a database or data warehouse, the 4 ta usually needs to be standardized, checked, and sometimes transformed to some common format and content. This allows faster search and retrieval, and serves as a check of data integrity. The rules that define the cleaning/trans-formation process are often termed "business rules," and in the case of chemical data, they may include checking and modification of chemical structures. 

ChoSY,ChaiJC,ParkSJetal (2013)EPITRANS a database that integrates epigenome and tran-scriptome data. Mol Cells 36 472-475... 

---