AnIML is based on XML technology and thus requires implementers to have some working knowledge of reading and writing XML. As XML is a very popular technology and actively being developed and used there exists a wealth of information online to get users up to speed.
XML can be read and written by all of the current software technologies such as:
Practically, to be able to collect, save, visualize and archive AnIML files in the laboratory, readers, writers, and viewers need to be written for the AnIML specification by different groups and made available for free or as part of commercial software. Groups include
- Instrument Vendors
AnIML will be usable for every type of analytical instrumentation so instrument companies will need to integrate export functionality in their software.
- Analytical Laboratories
Laboratories will want to move toward storing analytical data in the AnIML format because of its features for data integrity and archiving.
- LIMS Vendors
LIMS systems currently in use in larger laboratories will need to integrate options to store and process AnIML files as companies move to consolidate and archive instrument data in one format.
- Government Agencies
Reporting of instrument data to different government agencies is likely to increase significantly as requirements become more stringent and AnIML is likely going to be the format of choice. Using AnIML will not only make compatibility significantly easier but will also make data more easily available for download.
Scientists currently deal with a variety instruments producing multiple file formats and have to keep software to read these data files - forever! Conversion or export to AnIML will significantly improve this situation as generic viewers will always be available for an open standard. In addition, AnIML can be used to store data from research equipment where software can be designed specifically to write the specification. In fact, researchers can write their own technique definitions for in-house built research equipment if they wish (although these would not be an official part of the AnIML specification).
Metadata and Crosswalking
So, you have a lot data in your LIMS system or as JCAMP files. How do you move data to AnIML? In general the following steps are needed for conversion of legacy data
- Identification of the number of file types requiring conversion
- Development of a crosswalk (see this presentation) of the metadata and raw data in each format to the AnIML specification, including how to handle missing data elements.
- Identification of missing, complementary, or relational metadata that needs to be included in the converted AnIML file (e.g. sample, user, instrument properties)
- Consolidation of all the metadata need to go into an AnIML in an intermediate/temporary format (e.g. a specially design database table)
- Development of software routines to parse the metadata into the AnIML file format
- Post processing file audits to verify that the data has been correctly converted
While this process may take some time, it will result in a significantly more useful dataset. Lessons learned in the conversion can be fed back into the laboratory process to improve the quality/extent of metadata and data that is collected.