NEED FOR COLLABORATIVE TECHNOLOGIES IN DRUG DISCOVERY - Collaborative Computational Technologies for Biomedical Research - page 9

Figure 1.3 Emergence of a selectively integrated drug discovery and development

model.

For example, for every pharmaceutical company, there may be two or three
chemistry synthesis partners. These partners would likely have their own inter-
nal systems for tracking reagents, recording experiments, and registering novel
compounds. Since the synthesis is performed on behalf of the pharmaceutical
client, a majority of the data from the experiment, from reaction yields to
analytical data, must be transmitted to the client along with the synthesized
compound in a vial. The challenge is that since the pharmaceutical client has
developed mature internal processes, and the synthesis partner has its own
internal processes, there is a high likelihood that the processes—and the
related IT systems—are different in nature. This leads to the use of different
metadata, different vocabularies, and different quality control on the data
capture. When an instance of a novel compound is synthesized, the outsource
partner may call it a “batch” but the pharmaceutical client may call it a “lot”.
Also, some compound registration systems assign a different identifi er for dif-
ferent salt forms of the compound. One company may handle this by using a
suffi x of the compound identifi er ( ),

whereas another company may simply assign a completely different base
compound identifi er to the different salt form. Both of these are legitimate
taxonomies to register and identify compounds and their salt forms. The dif-
fi culty comes when one company attempts to export its registration data and
transmit that to the other company. Reconciling the differences in the seman-
tics and vocabularies of different compound registration systems can be a
tedious, error-prone, and often irreconcilable task. Often this reconciliation
involves compound registrars and synthetic chemists (and possibly lawyers)
from both parties. If the need to transmit compound registration data between
business partners was a unique event, then perhaps a manual reconciliation
process would suffi ce. However, since every pharmaceutical company has
several synthesis outsourcing partners, and every synthesis CRO has several
pharmaceutical clients, this metadata-confl ict and reconciliation process is

< compound identifi er > - < salt form >