Use of Molecular Target Measurements as a COMPARE seed

COMPARE has been successful in identifying compounds from the screening database with a mechanism of action similar to that of a known seed compound. It is reasonable to assume that this success is due to the varying importance of the target pathway for the growth of the various cell lines in the panel. Cell lines that rely heavily on a particular pathway will be very sensitive to inhibition of that pathway and in cell lines where that pathway is absent, drugs which target that pathway will have little or no effect on cell growth. Thus the characteristic pattern that enables COMPARE to group together drugs with a similar mechanism can be viewed as an indirect measure of the variabilty of the target pathway across the cell lines in the panel. This raises the question of whether it would be possible to make a direct measurement of a particular target across the cell line panel and use this pattern directly as a COMPARE seed to pick out drugs that act by attacking that particular target. It would then be potentially possible to discover drugs which attack this target without having any previous known examples and without knowing the structure of the target.

Targets which are important in tumor biology and thus potentially new targets for drug discovery have been identified. These molecules, which include oncogenes, suppressor genes, and growth factors/receptors, could be overexpressed, underexpressed, or mutated in cancer cells. The expression of a number of these targets has been measured in the tumor cell panel and the resulting pattern used a COMPARE seed. The screening data was searched for compounds which most closely correlated with target expression in both a positive or negative sense, i.e., which compounds were most sensitive in cell lines that overexpressed or underexpressed the targets. For targets which were mutated in tumor cells, data were expressed in binary mode ( the mutation is present in the particular cell line or it is not ).

A number of different studies completed to date have shown that using this approach COMPARE, coupled with laboratory verification, has been successful in identifying new agents which affect the molecular target. Other studies have identified agents which are selectively effective in lines altered in expression or mutation of the target. These agents should be especially valuable in studies designed to understand the role of these molecular targets in the tumor phenotype. Although this methodology is young and the need to verify any new correlations in the laboratory must be emphasised, these early successes have encouraged studies on a broad range of targets.

Return to main page