Since the complete genome sequence for COVID-19 was first published as NCBI GenBank MN908947 by Chinese scientists in early January of this year, COVID-19 genome sequences have quickly been made freely available in the NCBI GenBank, the Virological open access repository, and GISAID database. Since availability of this detailed data, the global response from researchers in academia and industry has rapidly accelerated. For example, according to the International Federation of Pharmaceutical Manufacturers and Associations (IFPMA), clinical trials for treatments and COVID-19 vaccines are already in their initial trial phases. Individual electronics companies are supporting these and other efforts directly or through research alliances to provide resources needed to accelerate discovery.
Yet, work in providing desperately needed support for COVID-19 research and clinical care is only beginning, and multiple industry and crowdsource efforts are underway to help mitigate the COVID-19 pandemic. We can’t hope to list every such effort – and more continual to appear daily in journals, open access sites arXiv, ChemRxiv and medRxiv and bioRxiv In addition, pop up efforts trend rapidly through social media to respond to immediate needs. Nevertheless, here’s a quick summary of a few activities that companies, coders and hardware developers can participate in to help build the COVID-19 knowledge base and equipment base.
Kaggle COVID-19 Open Research Dataset Challenge (CORD-19)
Kaggle is using its approach for competitive AI development to enlist its community of AI developers to provide researchers and healthcare workers with the ability to more effectively utilize the COVID-19 Open Research Dataset (CORD-19). CORD-19 is a freely available dataset containing tens of thousands of articles about COVID-19, SARS-CoV-2, and related coronaviruses, but finding answers can be difficult particularly with the explosive growth in related scientific papers. Kaggle’s CORD-19 challenge asks developers to use text and data mining methods on the CORD-19 dataset to provide answers to specific questions posed by scientists and healthcare workers. Kaggle provides the dataset and defined specific tasks that challenge participants need to complete satisfactorily through two evaluation rounds ending April 16 and June 16 to be eligible for a $1,000 prize. Earlier Kaggle COVID-19 competitions have focused on forecasting COVID-19 transmission and curating COVID-19 datasets.
XPRIZE Pandemic Alliance
XPRIZE has launched its Pandemic Alliance, which is designed to enhance visibility into emerging solutions and use of datasets and AI techniques to find new solutions. Partnering in this launch, Anthem is providing eligible users with access to its certified data set of longitudinal data on previous viral outbreaks. Other partners include Intel, Nvidia, IEEE, and gene sequencing equipment supplier Illumina, among others. One partner, the MIT Solve initiative for tech entrepreneurs, has recently launched its own 2020 Challenge: Health Security & Pandemics providing access to over $1M in funding.
IBM 2020 Call for Code Global Challenge
IBM is focusing its 2020 Call for Code Global Challenge on enhancing capabilities needed to assist in the global response to COVID-19. Although developers can pursue their own application focus, IBM provides starter kits in three different target areas: crisis communications, remote education and community collaboration. Initial submissions are due on April 27, 2020, and the top three solutions will be announced at the IBM Think online conference in early May. Following deployment and final submission, the global challenge winners will be announced in October 2020.
Folding@home background simulation processing
Improved understanding of COVID-19 at the molecular level is important for focusing efforts on the most promising avenues for prevention and cure. Complex biomolecules naturally fold and unfold to reveal or hide binding sites used by other molecules to effect some bioreaction. A clear understanding of the way the virus folds under different conditions is particularly vital under pandemic conditions to help “wet lab” researchers focus their efforts on the more likely candidate inhibitors and antagonists. To simulate COVID-19 folding, Folding@home is using a background distributed processing method used for years in crowdsourced distributed computing platforms such as SETI@Home. A related Folding@home github repo provides input files and generated datasets for Folding@home simulations.
Figure 1: Simulation of the spike COVID19 spike protein with three colors showing different
proteins that assemble together and showing how the end towards the top opens up
to reveal the binding site that engages with human cells. (Source: Folding@home)
Open source respiratory aid designs
As described more fully in an earlier article, multiple projects are working to implement low cost ventilator systems. Among these, an open-source github project describes how developers might convert a typical Continuous Positive Airway Pressure (CPAP) blower into a non-invasive pressure support Bilevel Positive Airway Pressure (BiPAP) device. The author, Johnny Lee, is quick to point out that such a device could at best relieve some level of respiratory discomfort, but it cannot replace a proper pulmonary ventilator. Indeed, medical ventilators are more than air pumps but need to address vital requirements related to pulmonary function. Johnny Lee does offer an open-source design for a low cost Powered Air Purifying Respirator (PAPR) backpack (see figure 2 below). PAPR equipment is typically used by healthcare workers, emergency responders, and those working in hazardous industrial environments. He has also compiled a list of related projects and resources for those interested in pursuing similar efforts.
Figure 2: DIY Powered Air Purifying Respirator (PAPR) backpack system
that can be created using off the shelf parts. (Source: Johnny Lee)
Finding embedded developer resources
Companies in a position to pivot their manufacturing lines to create healthcare equipment including pulmonary ventilators may themselves lacking hardware and software engineering expertise required to rapidly design the underlying electronic systems. CIS Electronics Engineering is providing its services as a free intermediary for companies that need engineers and for engineers able and willing to donate their expertise to such efforts. CIS Electronics Engineering provides a simple method for companies and engineers to participate in this effort.
Wide and growing availability
As noted earlier, the efforts described above only scratch the surface of the global response to the COVID-19 pandemic. Online service providers provide decision makers with a wide range of free tools and services such as Tableau’s Global COVID-19 Starter Workbook and Google’s Maps-Platform COVID-19 Developer Resource Hub. More broadly, the COVID-19 High Performance Computing Consortium provides eligible researchers with free compute time and resources weighing in at 366 petaflops, 64,092 nodes, 2,839,772 CPU cores, and 36,058 GPUs as of this writing.
If you are offering or using a relevant service that should be included here, please mention it in the comments below.