OTTM: an automated classification tool for translational drug discovery from omics data

Abstract

Omics data from clinical samples are the predominant source of target discovery and drug development. Typically, hundreds or thousands of differentially expressed genes or proteins can be identified from omics data. This scale of possibilities is overwhelming for target discovery and validation using biochemical or cellular experiments. Most of these proteins and genes have no corresponding drugs or even active compounds. Moreover, a proportion of them may have been previously reported as being relevant to the disease of interest. To facilitate translational drug discovery from omics data, we have developed a new classification tool named Omics and Text driven Translational Medicine (OTTM). This tool can markedly narrow the range of proteins or genes that merit further validation via drug availability assessment and literature mining. For the 4489 candidate proteins identified in our previous proteomics study, OTTM recommended 40 FDA-approved or clinical trial drugs. Of these, 15 are available commercially and were tested on hepatocellular carcinoma Hep-G2 cells. Two drugs-tafenoquine succinate (an FDA-approved antimalarial drug targeting CYC1) and branaplam (a Phase 3 clinical drug targeting SMN1 for the treatment of spinal muscular atrophy)-showed potent inhibitory activity against Hep-G2 cell viability, suggesting that CYC1 and SMN1 may be potential therapeutic target proteins for hepatocellular carcinoma. In summary, OTTM is an efficient classification tool that can accelerate the discovery of effective drugs and targets using thousands of candidate proteins identified from omics data. The online and local versions of OTTM are available at http://otter-simm.com/ottm.html.

Keywords: OTTM; hepatocellular carcinoma; literature mining; omics data; translational drug discovery.

Figure 1

Proteins or genes from omics data are classified into four categories by OTTM. OTTM can classify thousands of DEPs or genes identified from omics data into four categories. Category 1: proteins have corresponding drugs but have been reported as being relevant to the disease of interest. Category 2: proteins have corresponding drugs and have not been reported as being relevant to the disease of interest. Category 3: proteins have no corresponding drugs and have been reported as being relevant to the disease of interest. Category 4: proteins have no corresponding drugs and have not been reported as being relevant to the disease of interest. OTTM aims to identify Category 2 proteins or genes from among all the proteins or genes identified from omics data.

Figure 2

Schematic diagram of the principles involved in how OTTM classifies its target proteins via drug availability assessment. (A) Schematic illustration of how OTTM classifies target proteins from omics data. The yellow capsules with a capital ‘D’ represent target proteins that have corresponding drugs but have been previously reported as being relevant to the disease of interest. The red capsules with a capital ‘D’ represent target proteins that have corresponding drugs and have not been previously reported as being relevant to the disease of interest. The yellow and red circles represent target proteins without corresponding drugs. (B) From an initial 4489 DEPs, 645 target proteins with corresponding drugs were identified by OTTM. Of these, 88 (13.64%) are associated with FDA-approved drugs, 182 (28.22%) with drugs under clinical trial and 375 (58.14%) with drugs that have been discontinued. (C) Among the 88 target proteins associated with FDA-approved drugs, OTTM suggests that eight have not been reported as being relevant to hepatocellular carcinoma, whereas each of the other 80 is mentioned in at least one PubMed abstract that contains the designated keyword ‘hepatocellular carcinoma’. (D) Among the 182 target proteins associated with drugs under clinical trial, OTTM suggests that 15 have not been reported as being relevant to hepatocellular carcinoma, whereas each of the other 167 is mentioned in at least one PubMed abstract that contains the designated keyword ‘hepatocellular carcinoma’.

Figure 3

Drugs and target proteins recommended by OTTM for 4489 DEPs. (A) Target classification based on drug availability and literature mining in PubMed abstracts using ‘hepatocellular carcinoma’ as the designated keyword. OTTM suggested that 23 target proteins have not been reported as being relevant to hepatocellular carcinoma, including eight target proteins with FDA-approved drugs and 15 target proteins with drugs under clinical trial. Users can interactively click on each branch of this tree diagram to expand or collapse each category. (B) CALCRL is used as an example to illustrate the drug classification for a target protein. First, the drugs corresponding to a target protein are classified into FDA-approved and clinical trial categories. Then, these drugs are further classified into Reported and Not Reported categories based on the results of the literature mining using the designated keyword ‘hepatocellular carcinoma’. (C) Seven drugs corresponding to the target protein CALCRL were ranked by OTTM according to the number of PubMed abstracts containing their names. (D) From an initial 4489 candidate proteins, 23 drugs and 23 target proteins were recommended by OTTM for further experimental validation, including eight FDA-approved and 15 clinical trial drugs. One drug is recommended for each target protein.

Figure 4

Drugs and target proteins recommended by OTTM for 2951 interacting proteins. (A) The principle involved in how OTTM constructs the list of proteins that interact with the user-provided proteins according to PPI information obtained from the STRING database. The cyan circles with uppercase ‘U’ represent candidate proteins provided by users, whereas the yellow circles with uppercase ‘P’ represent interacting proteins. For the 4489 candidate proteins provided, 2951 interacting proteins were identified by OTTM and formed the protein list used for further target classification and drug recommendations. (B) Target classification by OTTM of the 2951 interacting proteins based on drug availability assessment. OTTM indicated that 79 interacting proteins have FDA-approved drugs, whereas 100 interacting proteins have drugs undergoing clinical trials. (C) From an initial 2951 interacting proteins, 17 drugs and target proteins were recommended by OTTM for further experimental validation, including nine FDA-approved and eight clinical trial drugs. One drug is recommended for each target protein.

Figure 5

Cell viability assay for the 15 commercially available drugs recommended by OTTM. Combined, OTTM recommended 40 drugs, including 23 for the 4489 candidate proteins and 17 for the 2951 interacting proteins. Among the 40 drugs recommended by OTTM, 15 were commercially available and were tested for their activity against the HepG2 hepatocellular carcinoma cell line. The target proteins of these 15 drugs were classified into two categories by source. DEP represents the DEPs identified from the omics data and PPI represents the interacting proteins according to existing PPI information from the STRING database.

Figure 6

General workflow of OTTM usage. The usage of OTTM includes four steps. Step 1: visit the OTTM web server for online and local usage. Step 2: for local usage, users must download the executable packages and necessary data. For online usage, users must upload a protein list and a configuration file. Step 3: inspect the OTTM-generated output results for recommended drugs. Step 4: purchase the available drugs recommended by OTTM and assess their efficacy in experiments.