Generally speaking, automation means lots of data. Without a data management system, the flood of data generated from automation equipment can render the whole approach useless as the challenge of carrying out countless experiments simply gets transferred into the task of sifting through countless result files.
For protein crystallization in particular, the primary challenge is image management. This is because of the large number of images taken over the course of a project together with the resolution requirements and the associated image files sizes. To address this problem it is common practice to store image files on RAID arrays and use a relational database to manage the image meta data (i.e. experiment ID, well ID, time stamps, and others).
Another area where data management affects automation is tracking of experiments. One choice is to assign plates to experiments or projects and track the experiment or project IDs through the life cycle of a protein crystallization experiments. Another method is to assign bar codes to plates and deep well block (or any other item handled by the automation equipment, like stocks for example). This has the one obvious advantage that bar codes can be easily read by robots, but the real benefit is that bar codes prevent manual manipulation which reduces errors.
All this of course makes only sense if screen information, inspection results, scoring information and all other relevant information is readily available via search and query functionality.
Click here to learn more about data management at Rigaku.com.
