Curation of Inhibitor-Target Data: Process and Impact on Pathway Analysis

Abstract

The past decade has seen a significant emergence in the availability and use of pathway analysis tools. The workflow that is supported by most of the pathway analysis tools is limited to either of the following:

• a. a network of genes based on the input data set, or

• b. the resultant network filtered down by a few criteria such as (but not limited to)

  • i. disease association of the genes in the network;

  • ii. targets known to be the target of one or more launched drugs;

  • iii. targets known to be the target of one or more compounds in clinical trials; and

  • iv. targets reasonably known to be potential candidate or clinical biomarkers.

Almost all the tools in use today are biased towards the biological side and contain little, if any, information on the chemical inhibitors associated with the components of a given biological network. The limitation resides as follows:

• The fact that the number of inhibitors that have been published or patented is probably several fold (probably greater than 10-fold) more than the number of published protein–protein interactions. Curation of such data is both expensive and time consuming and could impact ROI significantly.

• The non-standardization associated with protein and gene names makes mapping reasonably non-straightforward.

• The number of patented and published inhibitors across target classes increases by over a million per year. Therefore, keeping the databases current becomes a monumental problem.

• Modifications required in the product architectures to accommodate chemistry-related content.

GVK Bio has, over the past 7 years, curated the compound-target data that is necessary for the addition of such compound-centric workflows. This chapter focuses on identification, curation and utility of such data.