Journal: bioRxiv

Article Title: Enabling the hypothesis-driven prioritization of ligand candidates in big databases: Screenlamp and its application to GPCR inhibitor discovery for invasive species control

doi: 10.1101/249151

Figure Lengend Snippet: Summary of the tools provided or augmented by Screenlamp. (1) Filtering tasks that can be performed within Screenlamp to meet hypothesis-driven criteria and retrieve the structures of a subset of candidate molecules. (2) Once flexible conformers of the candidate database molecules have been sampled and overlaid with the reference molecule (for example, by using OMEGA and ROCS from OpenEye), Screenlamp can identify functional group matches in the pairwise overlays to discover functional group mimics of a reference molecule. (3) Based on the information that is available from the 3D overlays and functional group matching, as well as user-specified selection criterion, molecules are ranked for experimental testing.

Article Snippet: To prepare this database for our virtual screening pipeline, partial charges were added to the existing 3D structures of these molecules using OpenEye QUACPAC/molcharge with the AM1BCC force field, and low-energy conformations were generated using OpenEye OMEGA with default settings.

Techniques: Functional Assay, Selection

Journal: bioRxiv

Article Title: Enabling the hypothesis-driven prioritization of ligand candidates in big databases: Screenlamp and its application to GPCR inhibitor discovery for invasive species control

doi: 10.1101/249151

Figure Lengend Snippet: 3D structures of the 15 most active molecules from screening the ZINC drug-like dataset, the combinatorial analog dataset, CAS registry steroids, and the GPCR Ligand Library as described in the Methods section, Preparation of millions of drug-like molecules for ligand-based screening. The molecule structures are numbered from highest percent inhibition (#1) to lowest (#15). Accession codes in the CAS registry and ZINC databases are provided along with average percent inhibition values over two or more replicates. Hypothesis-based selection criteria are listed below the compound IDs, referencing the hypothesis descriptions given in the Results. The ZINC13057041 compound has been flagged as a potential PAIN (pan-assay interference compound) containing a functional group that leads to false-positive assay results via the server available at http://cbligand.org/PAINS/ .

Article Snippet: To prepare this database for our virtual screening pipeline, partial charges were added to the existing 3D structures of these molecules using OpenEye QUACPAC/molcharge with the AM1BCC force field, and low-energy conformations were generated using OpenEye OMEGA with default settings.

Techniques: Inhibition, Selection, Pan Assay, Functional Assay

Journal: bioRxiv

Article Title: Enabling the hypothesis-driven prioritization of ligand candidates in big databases: Screenlamp and its application to GPCR inhibitor discovery for invasive species control

doi: 10.1101/249151

Figure Lengend Snippet: (a) Heat map showing the functional group matches of the 15 most active and 15 least active molecules when overlaid with 3kPZS. The percent inhibition was computed as the average inhibition over two or more independent electro-olfactogram assays. Heat map cells containing 1’s indicate the presence of a match and 0’s indicate the absence of a match. (b) 3D representation of an energetically favorable 3kPZS conformer with functional group labels, to aid in interpreting the heat map x-axis labels. We have developed computational protocols to identify activity discriminants from such data; see and the example code on GitHub: https://github.com/psa-lab/predicting-activity-by-machine-learning .

Article Snippet: To prepare this database for our virtual screening pipeline, partial charges were added to the existing 3D structures of these molecules using OpenEye QUACPAC/molcharge with the AM1BCC force field, and low-energy conformations were generated using OpenEye OMEGA with default settings.

Techniques: Functional Assay, Inhibition, Activity Assay

Journal: bioRxiv

Article Title: Enabling the hypothesis-driven prioritization of ligand candidates in big databases: Screenlamp and its application to GPCR inhibitor discovery for invasive species control

doi: 10.1101/249151

Figure Lengend Snippet: (a) Enrichment graphs showing the percent inhibition as a function of chemical and volumetric similarity of candidate molecules to 3kPZS, based on inhibition assays for 299 molecules selected by structure-activity hypotheses. The overlay score threshold refers to the ROCS TanimotoCombo score, equally weighting volumetric and pharmacophore similarity. The different traces on the graphs compare the enrichment for several hypotheses relative to using 3D similarity (overlay score) alone. (b) Receiver operating characteristic curve (rate of retrieval of true positives vs. false positives), with triangles, square and circle symbols used to show the point on each curve corresponding to a given score threshold, for cases in which the overlay score alone was used to select candidate compounds, versus when the overlay score was augmented by increasingly selective steroid-based hypotheses, resulting in fewer molecules being tested but a greater enrichment in active compounds. The curves show the number of active molecules (defined as having at least 50 percent inhibition of 3kPZS in experimental assays) for different overlay thresholds.

Article Snippet: To prepare this database for our virtual screening pipeline, partial charges were added to the existing 3D structures of these molecules using OpenEye QUACPAC/molcharge with the AM1BCC force field, and low-energy conformations were generated using OpenEye OMEGA with default settings.

Techniques: Inhibition, Activity Assay