A deep learning framework for high-throughput mechanism-driven phenotype compound screening and its application to COVID-19 drug repurposing

Abstract

Phenotype-based compound screening has advantages over target-based drug discovery, but is unscalable and lacks understanding of mechanism. Chemical-induced gene expression profile provides a mechanistic signature of phenotypic response. However, the use of such data is limited by their sparseness, unreliability, and relatively low throughput. Few methods can perform phenotype-based de novo chemical compound screening. Here, we propose a mechanism-driven neural network-based method DeepCE, which utilizes graph neural network and multi-head attention mechanism to model chemical substructure-gene and gene-gene associations, for predicting the differential gene expression profile perturbed by de novo chemicals. Moreover, we propose a novel data augmentation method which extracts useful information from unreliable experiments in L1000 dataset. The experimental results show that DeepCE achieves superior performances to state-of-the-art methods. The effectiveness of gene expression profiles generated from DeepCE is further supported by comparing them with observed data for downstream classification tasks. To demonstrate the value of DeepCE, we apply it to drug repurposing of COVID-19, and generate novel lead compounds consistent with clinical evidence. Thus, DeepCE provides a potentially powerful framework for robust predictive modeling by utilizing noisy omics data and screening novel chemicals for the modulation of a systemic response to disease.

Predictive Modeling of Multiplex Chemical Phenomics for Novel Cells and Patients: Applied to Personalized Alzheimer’s Disease Drug Repurposing

“…The copyright holder for this preprint this version posted August 10, 2021. ; https://doi.org/10.1101/2021.08.09.455708 doi: bioRxiv preprint 22 novel chemical representation module in DeepCE, our model can satisfy the prediction of chemical phenomics under the circumstance of both novel cells/patients and novel drugs 35 . We demonstrated that the predicted chemical transcriptomics were informative to downstream applications.…”

“…We augmented our data using a modified procedure described elsewhere 35 . We performed the data augmentation procedure multiple times which is inspired by teacher-student model.…”

Artificial Intelligence in Surveillance, Diagnosis, Drug Discovery and Vaccine Development against COVID-19

“…Pham et al proposed DeepCE, a deep learning algorithm to repurpose drug compounds. The author demonstrated the application of DeepCE to predict potential leads for COVID-19 treatment [ 146 ]. In another study, an ML model was built to predict new indications for existing drugs and herbal compounds based on 1330 positive drug-disease associations though it was not directed against COVID-19 [ 147 ].…”

Nonsteroidal anti-inflammatory drugs and glucocorticoids in COVID-19

“…Remdesivir is the only antiviral drug approved or authorized for emergency use from several international drug agencies, but not by the World Health Organization (WHO), to treat hospitalized COVID-19 patients ( https://www.covid19treatmentguidelines.nih.gov/antiviral-therapy/remdesivir/ ; https://www.ema.europa.eu/en/news/ema-provides-recommendations-compassionate-use-remdesivir-covid-19 ; https://www.tga.gov.au/media-release/australias-first-covid-treatment-approved ; https://www.who.int/news-room/feature-stories/detail/who-recommends-against-the-use-of-remdesivir-in-covid-19-patients ). While the potential clinical efficacy of molecules emerging from high throughput screens is explored ( Pham et al, 2021 ), the use of dexamethasone in patients with severe disease has been established with the Randomised Evaluation of COVID-19 Therapy (RECOVERY) trial ( Horby et al, 2021 ). More controversial is the use of glucocorticoids in patients who are not on supplemental oxygen in the Intensive Care Unit and the use of nonsteroidal anti-inflammatory drugs (NSAID)s to quell the immune response at any stage of the disease.…”