Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
80 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
7 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Repurpose Open Data to Discover Therapeutics for COVID-19 using Deep Learning (2005.10831v1)

Published 21 May 2020 in q-bio.QM, cs.LG, and stat.ML

Abstract: There have been more than 850,000 confirmed cases and over 48,000 deaths from the human coronavirus disease 2019 (COVID-19) pandemic, caused by novel severe acute respiratory syndrome coronavirus (SARS-CoV-2), in the United States alone. However, there are currently no proven effective medications against COVID-19. Drug repurposing offers a promising way for the development of prevention and treatment strategies for COVID-19. This study reports an integrative, network-based deep learning methodology to identify repurposable drugs for COVID-19 (termed CoV-KGE). Specifically, we built a comprehensive knowledge graph that includes 15 million edges across 39 types of relationships connecting drugs, diseases, genes, pathways, and expressions, from a large scientific corpus of 24 million PubMed publications. Using Amazon AWS computing resources, we identified 41 repurposable drugs (including indomethacin, toremifene and niclosamide) whose therapeutic association with COVID-19 were validated by transcriptomic and proteomic data in SARS-CoV-2 infected human cells and data from ongoing clinical trials. While this study, by no means recommends specific drugs, it demonstrates a powerful deep learning methodology to prioritize existing drugs for further investigation, which holds the potential of accelerating therapeutic development for COVID-19.

Repurposing Open Data with Deep Learning for COVID-19 Therapeutic Discovery

The paper "Repurpose Open Data to Discover Therapeutics for COVID-19 using Deep Learning" introduces a sophisticated approach to drug repurposing in the context of the COVID-19 pandemic, leveraging advanced network-based deep learning techniques. At its core, the paper presents CoV-KGE, an integrative framework designed to systematically identify existing drugs with potential therapeutic effects against SARS-CoV-2, the virus responsible for COVID-19.

Methodology and Implementation

The authors constructed a comprehensive knowledge graph encompassing 15 million edges across 39 relationship types connecting crucial biomedical elements such as drugs, genes, diseases, and pathways. This was generated by mining data from an extensive corpus of 24 million publications in PubMed. Utilizing the DGL-KE tool developed by Amazon's AWS AI Lab, the team employed the RotatE model for learning low-dimensional embeddings that informed their analysis. The embeddings represent entities and relationships within the graph in a vector space, facilitating the prediction of potential drug candidates through relational rotations.

The methodology was applied to prioritize candidate drugs by computing vectors associated with their therapeutic potential against COVID-19. This process highlighted the top-100 candidate drugs, which were subsequently validated using both transcriptomic and proteomic data from SARS-CoV-2 infected cells and corroborative data from ongoing clinical trials.

Results and Validation

A significant outcome of this paper is the identification of 41 repurposable drugs, among which indomethacin, toremifene, and niclosamide were noted for their potential efficacy against COVID-19, supported by existing clinical and in vitro evidence. The reported area under the curve (AUROC) value of 0.85 underscores the high predictive performance of the CoV-KGE system in drug repurposing for COVID-19.

These findings were further validated using enrichment analysis of drug-gene signatures and SARS-CoV induced transcriptomic datasets, reinforcing the method's robustness. Notably, this work not only confirms existing candidates that are presently under investigation in clinical trials but also suggests newer representatives from diverse drug classes including anti-inflammatory agents, selective estrogen receptor modulators, and antiparasitics.

Implications and Future Directions

The implications of this paper are multifaceted. Practically, it provides a rapid, cost-effective strategy to pivot existing drug compounds towards addressing COVID-19 treatment challenges, potentially accelerating therapeutic development in this critical public health crisis. Theoretically, this framework offers insight into how network-based methodologies can be deployed to exploit vast biomedical data resources for emergent infectious diseases.

Looking forward, the CoV-KGE approach could be generalized beyond COVID-19 to facilitate expedited drug discovery for other complex diseases and novel pathogens. As the field of AI in drug discovery continues to mature, harnessing such integrative platforms may provide significant leverage in addressing wide-ranging biomedical challenges.

This paper epitomizes the intersection of AI, computational biology, and pharmacology, presenting a solid blueprint for future explorations into data-driven biomedical innovation.

User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (9)
  1. Xiangxiang Zeng (28 papers)
  2. Xiang Song (34 papers)
  3. Tengfei Ma (73 papers)
  4. Xiaoqin Pan (3 papers)
  5. Yadi Zhou (2 papers)
  6. Yuan Hou (3 papers)
  7. Zheng Zhang (488 papers)
  8. George Karypis (110 papers)
  9. Feixiong Cheng (5 papers)
Citations (187)