key: cord-0886512-9xgrthg7
authors: Ahuja, Abhimanyu S.; Reddy, Vineet Pasam; Marques, Oge
title: Artificial Intelligence and COVID-19: A Multidisciplinary Approach
date: 2020-05-27
journal: Integr Med Res
DOI: 10.1016/j.imr.2020.100434
sha: da5e1d82774ceef4a52721eacf4963ae97fb0a78
doc_id: 886512
cord_uid: 9xgrthg7

The COVID-19 pandemic is taking a colossal toll in human suffering and lives. A significant amount of new scientific research and data sharing is underway due to the pandemic which is still rapidly spreading. There is now a growing amount of coronavirus related datasets as well as published papers that must be leveraged along with artificial intelligence (AI) to fight this pandemic by driving news approaches to drug discovery, vaccine development, and public awareness. AI can be used to mine this avalanche of new data and papers to extract new insights by cross-referencing papers and searching for patterns that AI algorithms could help discover new possible treatments or help in vaccine development. Drug discovery is not a trivial task and AI technologies like deep learning can help accelerate this process by helping predict which existing drugs, or brand-new drug-like molecules could treat COVID-19. AI techniques can also help disseminate vital information across the globe and reduce the spread of false information about COVID-19. The positive power and potential of AI must be harnessed in the fight to slow the spread of COVID-19 in order to save lives and limit the economic havoc due to this horrific disease.

combat this pandemic. As requested by the White House Office of Science and Technology Policy, the COVID-19 Open Research Dataset has recently been released and contains over 29,000 machine learning (ML) readable articles concerning the novel coronavirus [1] . The White House has issued a call to action for the nation's AI experts to help answer key questions related to COVID-19 [1] . We believe that AI can be harnessed to combat this deadly pandemic and spur innovation in four key areas: drug discovery, vaccine development, public communications, and integrative medicine.

With the spread of showing no signs of slowing and its relatively high mortality rate when compared to other viral bases illnesses such as influenza, it is essential to develop vaccines and antiviral medications against SARS-CoV-2. In 2019, the National Institute of Allergy and Infectious Diseases sponsored the first U.S. clinical trial on an AI-based flu vaccine [2] . The vaccine was developed by scientists at Flinders University first using an AI program they created called synthetic chemist which generated trillions of synthetic compounds [2, 3] . The group then used another AI program called Search Algorithm for Ligands (SAM) which uses AI capabilities to sift through the trillions of compounds and determine which would be good candidates as vaccine adjuvants [2, 3] . This approach can shorten the vaccine development process by several years. While a vaccine for SARS-CoV-2 has already entered human trials, the efficacy is unknown. An AI-based approach could help screen compounds to be used as potential adjuvants for the SARS-CoV-2 vaccine, along with screening new compounds based on modeling of potential mutations to the novel coronavirus. This will help us be prepared to develop vaccines as the virus potentially mutates.

Furthermore, AI can aid in both screening currently available drugs and accelerating the process of antiviral development to help treat COVID-19. AI algorithms could be trained using the CORD-19 dataset and the resulting model could then be used to screen for existing drugs that potentially demonstrate efficacy in the treatment of COVID-19. This screening approach could be used, for example, to identify generic medications that may be effective against COVID-19. Regarding future drug design of SARS-CoV-2 antivirals, a team has already begun using the AIequipped IBM supercomputer SUMMIT to be used in the treatment of SARS-CoV-2 [4] . The team is using a computational model of the spike (S-protein) of SARS-CoV-2 to model its interaction with the human ACE2 receptor [4] . They are using SUMMIT to enact an ensemble docking virtual high-throughput screening campaign and to identify small-molecules which can either bind to the protein-receptor complex or the protein alone [4] . In doing so, small-molecules that have the ability to limit viral recognition of host cells or disrupt host-virus interactions can be rapidly identified [4] . Using this technique, trillions of small molecules can be rapidly modeled and consecutively tested.

While AI may assist in the discovery of novel drugs and vaccines, it can also help ease the stress placed on medical hotlines. Call centers nationwide that are dealing with the COVID-19 pandemic are notoriously understaffed. The Washington Department of Health (DOH) repurposed their Poison Center hotline, originally consisting of only four individuals, to advise concerned citizens [5] . In public health crises, it is difficult for state departments to allocate resources in order to scale up operations and reliably service callers. Though the Washington DOH is slowly increasing staffing and contacting other vendors, the hotline wait times still reach up to 38 minutes [5] . Furthermore, limitations in technology allowed the hotline to only keep a maximum of ten calls on hold [5] . In public health crises, it is critically important to ease public tension. With the unprecedented virality of COVID-19, call centers across the world are likely to be overtaxed, and people who test positive for COVID-19 may not receive proper medical guidance to manage the risk of COVID-19 and reduce its transmission. Artificial intelligence has the potential to revolutionize public communications and deliver alternative methods to dissipate public information.

AI-powered chatbots have been used with success in clinical scenarios and can advise many more people than a manned call center. Bibault et al. studied the quality of responses provided to patients with breast cancer from a team of physicians and an AI-powered chatbot [6] . The team reports noninferiority between the patients' rating of the quality of responses of the chatbot versus the physicians [6] . Greer et al. reports on Vivibot, a chatbot which acts as a psychosocial intervention for young adults who are cancer survivors [7] . This chatbot was advertised in-part through Facebook and deployed in Facebook Messenger. Rather than setting up hotlines, state departments could deploy chatbots through media services such as Facebook Messenger in order to quell panic and provide clinical information to the public [7] . This suggests that chatbots could be widely disseminated and aid more people than conventional call centers.

Thus far no aspects of integrative medicine have specifically been validated for the management of COVID-19 in randomized controlled trials [8] . However, there certainly is interest in the use of integrative medicine for the management of COVID-19. Currently, it is known the virulence and pathogenicity is associated with viral activation of cytoplasmic NLRP3 inflammasome [8] . This inflammasome in upregulated NFkb macrophages and Th1 cells results in proinflammatory cytokines (IL-1B and IL-18) [8] . These cytokines help mediate the pathogenic inflammation involved with much of the virulence and symptoms of COVID-19. This knowledge has enabled many potential integrative medicine considerations for risk reduction of COVID-19. This includes adequate sleep, stress management, zinc, Vegetables/ fruits +/-isolated Flavonoids, curcumin, melatonin, and Sambucus nigra (Elderberry) [8] . As more is learned about pathogenicity of SARS-CoV-2, an AI based approach may provide great benefit in screening possible integrative medicine techniques for the risk reduction of COVID-19. AI based screening methods could be used to specifically screen for integrative options that can help mediate the inflammatory responses to SARS-CoV-2 infection. Such an AI based approach can lead to a far more rapid deployment of integrative medicine techniques in the fight against COVID-19.

In the following weeks, we will likely see a colossal uptick of COVID-19 cases, both within the United States and globally. This pandemic arises due to the substantial virality of SARS-CoV-2, and locations of major sources of outbreak are facing a host of ground problems including a shortage of beds, ventilators, masks, gowns, and medical staff, amongst others. Though the situation may appear grim, it is important to draw attention to AI technologies that can ameliorate some of the most pressing and timely problems that this pandemic has created. Leveraging the power of AI, trillions of compounds can be screened in a remarkably short amount of time allowing for the rapid identification of drug and vaccine candidates. AI will certainly not be the only factor in the development of a successful antiviral drug or vaccine against SARS-CoV-2. Nevertheless, it is important for the scientific community to realize that we have powerful AI technology which can at the very least speed up the process. In this time of need, it is essential that pharmaceutical companies and basic science laboratories cooperate with industry leaders such as Google and IBM that have powerful AI capabilities. With COVID-19 spreading in our local communities, it is time to utilize our AI capabilities in the race to develop an effective vaccine and antivirals.

Frightened citizens are looking to receive guidance in these worrisome times and frequently turn to call centers for advice. Unfortunately, these call centers are limited by both staff and technological limitations. The amount of callers that a hotline can service is relatively small, and callers are placed on lengthy wait lists. This cluster of conundrums contributes to a rising sense of panic in the general population. The creation of chatbots that can be widely disseminated to educate masses of people across the globe can reduce panic and markedly curtail the spread of false information regarding COVID-19. People could ask for advice from state licensed or approved chatbots which can relay truthful and reliable information. These chatbots have already been built in popular messaging platforms that many young people already use, such as Facebook Messenger. People who are receiving false information on these platforms can be redirected to a state approved chatbot to receive effective and valid information. The presence of chatbots on familiar interfaces increases approachability and helps fight the misinformation also spreading in these applications.

Call to Action to the Tech Community on New Machine Readable COVID-19 Dataset. The White House

Identifier NCT03945825, Flublok or Fluzone With Advax-CpG55.2 or AF03

AI develops 'turbocharged' flu vaccine: A clinical trial will soon be underway in the U.S. to test the efficacy of an enhanced flu vaccine developed by an artificial intelligence system at Flinders University in South Australia. Becker's Hospital Review

Repurposing Therapeutics for COVID-19: Supercomputer-Based Docking to the SARS-CoV-2 Viral Spike Protein and Viral Spike Protein-Human ACE2 Interface. ChemRxiv

Coronavirus call centers stumble in Washington state: glitches, lack of staff, contradicting messages. The Seattle Times

A Chatbot Versus Physicians to Provide Information for Patients With Breast Cancer: Blind, Randomized Controlled Noninferiority Trial

Use of the Chatbot "Vivibot" to Deliver Positive Psychology Skills and Promote Well-Being Among Young People After Cancer Treatment: Randomized Controlled Feasibility Trial

Integrative considerations during the COVID-19 pandemic

This research has been approved by ______________ This research did not require an ethical approval as it does not involve any human or animal experiment. Did not require

The data will be made available upon request.

The data cannot be made available due to legal or ethical reasons as it contains sensitive information. The data generated and analyzed for this study are included in the supplementary material. The data has been deposited in the external database

Nothing to Disclose for any author.No Conflict of Interest for any author.J o u r n a l P r e -p r o o f