IDchase: Mitigating identifier migration trap in biological databases

A convenient mechanism to refer to large biological objects such as sequences, structures and networks is the use of identifiers or handles, commonly called IDs. IDs function as a unique place holder in an application for objects too large to be of immediate use in a table which is retrieved from a secondary archive when needed. Usually, applications use IDs of objects managed by remote databases that the applications do not have any control over such as GenBank, EMBL and UCSC. Unfortunately, IDs are generally not unique and frequently change as the objects they refer to change. Consequently, applications built using such IDs need to adapt by monitoring possible ID migration occurring in databases they do not control, or risk producing inconsistent, or out of date results, or even face loss of functionality. In this paper, we develop a wrapper based approach to recognizing ID migration in secondary databases, mapping obsolete IDs to valid new IDs, and updating databases to restore their intended functionality. We present our technique in detail using an example involving NCBI RefSeq as primary, and OCPAT as secondary databases. Based on the proposed technique, we introduce a new wrapper like tool, called IDChase, to address the ID migration problem in biological databases and as a general platform.