Virtual Affinity Fingerprints for Target Fishing: A New Application of Drug Profile Matching

Abstract

ABSTRACT:We recently introduced drug profile matching 12 (DPM), a novel virtual affinity fingerprinting bioactivity pre-13 diction method. DPM is based on the docking profiles of ca. 30 Finding compounds for a given target is a common computa-31 tional task in a conventional medicinal chemistry program. However, 32 by means of increasingly available bioactivity data, this 33 approach can be reversed to finding targets for compounds. 34 In silico target fishing 1 is an emerging field that aims at pre-35 dicting biological targets of molecules based on their chemical 36 structure. The rise of this area is in connection with that of 37 polypharmacology, 2,3 which posits that drugs act on multiple 38 targets in contrast with the traditional one drug−one target 39 paradigm. As a consequence, it is likely to discover new targets 40 even for well-known drugs. 41Many in silico target prediction tools have been developed, 42 and they were summarized by a recent review. 4 As it is common 43 for drug development methods, target prediction tools can also 44 be divided into two main groups: ligand-based and structure-45 based approaches. 46Similarity search is often used among the ligand-based methods. 47 The most common question that arises in case of similarity 48 based virtual screening is the description of molecular structure. 49 No universal solution seems to exist for this problem, 5 as the 50 best representation used to characterize the molecules depends 51 on the studied activity classes. Therefore, it is important to 52 combine several methods for a given task, e.g. by applying data 53 fusion techniques. 6 An approach that generates off-target profiles 54 of drugs based on their 3D similarity has just been reported, 55 and some of its predictions were proved by a literature survey.

Identification of PPARγ ligands with One-dimensional Drug Profile Matching

“…The starting hypothesis behind DPM is that similar affinity fingerprints refer to similar bioactivity properties (ie, targets and medical effects). We found that there is a strong relationship between the 177 medical effects studied and 77 target categories with the virtual affinity fingerprints of the drugs 3. Receiver operating characteristic analysis and cross-validations proved that the method is robust, and the affinity fingerprints provide enough information to recreate the effect profiles of drugs.…”

“…This method uses virtual affinity fingerprints of small molecule compounds to predict their medical effects and targets 2,3. In DPM, the affinity fingerprint is called IP, which consists of a set of docking scores of a compound against the binding sites of a predefined nontarget protein set.…”

“…In our earlier work, we found that at least ten members are required for a group to reliably perform multivariate statistical methods of DPM;2 with under ten members, an alternative evaluation method is required, like oDPM. Since the detailed mechanism of DPM has been introduced earlier,3 we will only summarize its most important features and point to the differences between oDPM and DPM.…”

Prediction of Protein−compound Binding Energies from Known Activity Data: Docking‐score‐based Method and its Applications

“…The quantitative structure−activity relationship (QSAR) approach is a useful tool for optimizing leads and predicting target/off‐target activities and toxicity. QSAR‐based affinity predictions are useful for the general drug development process, including the repositioning (repurposing) of already approved drugs, poly‐pharmacology, and the prediction of drug−drug interactions . The recent accumulation of protein−compound affinity data in public repositories, such as the PubChem and ChEMBL projects, has enabled us to carry out proteome‐wide target/off‐target predictions ,.…”

“…QSAR-based affinity predictions are useful for the general drug development process, including the repositioning (repurposing) of already approved drugs, poly-pharmacology, and the prediction of drugÀdrug interactions. [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15] The recent accumulation of proteinÀcompound affinity data in public repositories, such as the PubChem and ChEMBL projects, has enabled us to carry out proteome-wide target/off-target predictions. [16,17] These predictions are based on QSAR models for multiple proteins, just as in conventional computer-aided drug design and virtual screening.…”