Genomic-scale prioritization of drug targets: the TDR Targets database

Abstract

The increasing availability of genomic data for pathogens that cause tropical diseases has created new opportunities for drug discovery and development. However, if the potential of such data is to be fully exploited, the data must be effectively integrated and be easy to interrogate. Here, we discuss the development of the TDR Targets database (http://tdrtargets.org), which encompasses extensive genetic, biochemical and pharmacological data related to tropical disease pathogens, as well as computationally predicted druggability for potential targets and compound desirability information. By allowing the integration and weighting of this information, this database aims to facilitate the identification and prioritization of candidate drug targets for pathogens.

Figure 1. Searching the TDR Targets database

This figure shows the appearance of the search page and how different searches can be combined to yield lists of potential drug targets. First, the search can be restricted to a given species or groups of species; in this case Plasmodium falciparum is selected (top left panel). Next, one can query on function-based information. For example, selecting Enzyme under the Functional classific category of the Name/Annotation panel (top right) leads to a list of 862 genes identified in P. falciparum as encoding enzymes (bottom right). Similarly, the Structures panel below allows identification of the 4,829 genes for which there are crystal structures and/or high-quality structural models. Selecting within the “Phylogenetic distribution” panel (middle right) for orthologs present in Plasmodium vivax but not in humans or mice leads to 2,836 genes conserved in Plasmodium and absent in mammals. Very few genes have been validated as essential in P. falciparum, but an Essentiality search shows 1,162 P. falciparum genes with essential orthologs in at least one model organism – Caenorhabditis elegans, Escherichia coli, Mycobacterium tuberculosis, and Saccharomyces cerevisiae, and are therefore plausibly essential in P. falciparum. Each of these independent searches is stored in the History page of the database (bottom right). Registered users' queries are automatically saved for later sessions. The intersection of all four queries shown here have identified 25 potential drug targets for P. falciparum. These P. falciparum genes are potentially selective drug targets (they lack orthologous targets in mammals) that might allow for structure-based design of drugs (using available structures or models), are probably assayable (all are enzymes) and might be essential for viability (based on an ortholog being essential in some model organism). There are many promising drug targets in P. falciparum among this list (circled in red), supporting this approach as one means to prioritize promising targets.

Figure 2. Ranking of Mycobacterium tuberculosis targets using the TDR Targets database

This Venn diagram demonstrates the gene-ranking functionality of TDR Targets database. This example shows the weighted union of queries to yield a list of ranked genes from M. tuberculosis. The selected criteria, the arbitrary weights assigned to each criterion and the total number of genes returned for each query (bold font) are indicated in the diagram. Cumulative weights were assigned to genes intersecting multiple queries by summing the total weight of individual queries. For instance, the 8 genes that intersect all three queries – Query 1: enzymes (weighted 50) + having structure or structure model (weighted 20); Query 2: no human orthologue (weighted 30) + druggable (weighted 60); Query 3: essential in any species (weighted 50) + expressed in dormancy (weighted 30) – get the highest cumulative weight of 240. Table 2 shows a subset of the 1,563 total genes selected by the above criteria combination, showing the eight highest ranked genes at the top. Note that many of these highly ranked genes code for cytochrome P450 enzymes, which do of course have homologues in humans. However, they are selected in Query 2 (as well as the other queries) because they are sufficiently distinct from their human counterparts that they cluster into different ortholog groups. This suggests potential target selectivity.