Dataset Information

Modeling Small-Molecule Reactivity Identifies Promiscuous Bioactive Compounds.

ABSTRACT: Scientists rely on high-throughput screening tools to identify promising small-molecule compounds for the development of biochemical probes and drugs. This study focuses on the identification of promiscuous bioactive compounds, which are compounds that appear active in many high-throughput screening experiments against diverse targets but are often false-positives which may not be easily developed into successful probes. These compounds can exhibit bioactivity due to nonspecific, intractable mechanisms of action and/or by interference with specific assay technology readouts. Such "frequent hitters" are now commonly identified using substructure filters, including pan assay interference compounds (PAINS). Herein, we show that mechanistic modeling of small-molecule reactivity using deep learning can improve upon PAINS filters when modeling promiscuous bioactivity in PubChem assays. Without training on high-throughput screening data, a deep learning model of small-molecule reactivity achieves a sensitivity and specificity of 18.5% and 95.5%, respectively, in identifying promiscuous bioactive compounds. This performance is similar to PAINS filters, which achieve a sensitivity of 20.3% at the same specificity. Importantly, such reactivity modeling is complementary to PAINS filters. When PAINS filters and reactivity models are combined, the resulting model outperforms either method alone, achieving a sensitivity of 24% at the same specificity. However, as a probabilistic model, the sensitivity and specificity of the deep learning model can be tuned by adjusting the threshold. Moreover, for a subset of PAINS filters, this reactivity model can help discriminate between promiscuous and nonpromiscuous bioactive compounds even among compounds matching those filters. Critically, the reactivity model provides mechanistic hypotheses for assay interference by predicting the precise atoms involved in compound reactivity. Overall, our analysis suggests that deep learning approaches to modeling promiscuous compound bioactivity may provide a complementary approach to current methods for identifying promiscuous compounds.

SUBMITTER: Matlock MK

PROVIDER: S-EPMC8716318 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Parkinson's disease (PD) is the second most common neurodegenerative disorder with prominent dopamine (DA) neuron degeneration. PD affects millions of people worldwide, but currently available therapies are limited to temporary relief of symptoms. As an effort to discover disease-modifying therapeutics, we have conducted a screen of 1,403 bioactive small molecule compounds using an in vivo whole organism screening assay in transgenic larval zebrafish. The transgenic model expresses the bacterial enzyme nitroreductase (NTR) driven by the tyrosine hydroxylase (th) promotor. NTR converts the commonly used antibiotic pro-drug metronidazole (MTZ) to the toxic nitroso radical form to induce DA neuronal loss. 57 compounds were identified with a brain health score (BHS) that was significantly improved compared to the MTZ treatment alone after FDR adjustment (padj<0.05). Independently, we curated the high throughput screening (HTS) data by annotating each compound with pharmaceutical classification, known mechanism of action, indication, IC50, and target. Using the Reactome database, we performed pathway analysis, which uncovered previously unknown pathways in addition to validating previously known pathways associated with PD. Non-topology-based pathway analysis of the screening data further identified apoptosis, estrogen hormone, dipeptidyl-peptidase 4, and opioid receptor Mu1 to be potentially significant pathways and targets involved in neuroprotection. A total of 12 compounds were examined with a secondary assay that imaged DA neurons before and after compound treatment. The z'-factor of this secondary assay was determined to be 0.58, suggesting it is an excellent assay for screening. Etodolac, nepafenac, aloperine, protionamide, and olmesartan showed significant neuroprotection and was also validated by blinded manual DA neuronal counting. To determine whether these compounds are broadly relevant for neuroprotection, we tested them on a conduritol-b-epoxide (CBE)-induced Gaucher disease (GD) model, in which the activity of glucocerebrosidase (GBA), a commonly known genetic risk factor for PD, was inhibited. Aloperine, olmesartan, and nepafenac showed significant protection of DA neurons in this assay. Together, this work, which combines high content whole organism in vivo imaging-based screen and bioinformatic pathway analysis of the screening dataset, delineates a previously uncharted approach for identifying hit-to-lead candidates and for implicating previously unknown pathways and targets involved in DA neuron protection.

Dataset Information

Modeling Small-Molecule Reactivity Identifies Promiscuous Bioactive Compounds.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets