Open source tools are uniquely positioned to help combat the ongoing COVID-19 pandemic through their adaptable and collaborative nature. NumFOCUS sponsored and affiliated projects are being used on a global scale to meet the needs of researchers and data scientists. Our projects are being used in groundbreaking scientific efforts to create response models, visualize and analyze patient data, monitor the effects of lockdown, and simulate hospital capacities for local governments.
conda-forge offers 2500 scientific tools to the research community and provides access to huge computing resources around the globe, via a web-interface and accessibility for researchers worldwide. Galaxy is just one of the projects that heavily relies on conda-forge.
“Our Galaxy server has 2000+ conda environments, in each of them is the power of the conda-forge community! We schedule 500,000 jobs a month and all of them are using conda-forge packages.”
“[Galaxy and conda-forge are used] To develop a new drug against the Virus, this work analyzed over 40,000 compounds considered to be likely to bind to one of the SARS-CoV-2 proteases, which were chosen based on recently published X-ray crystal structures, and identified 500 high scoring compounds.”
– Björn Grüning (founding member of Galaxy)
Some of Galaxy and conda-forge’s applications included:
- Data Analysis through 16 COVID-19 workflows
- SARS-CoV-2 Dedicated Training Material and Webinars
- Compute Resource Provisions
- Hackathons and Co-Fests
Econ-Ark is impacting economic research surrounding COVID-19. Their pandemic dashboard has been a vital resource in producing papers such as “Modeling the Consumption Response to the CARES Act”. Read the abstract below:
“To predict the eﬀects of the 2020 U.S. ‘CARES’ act on consumption, we extend a model that matches responses of households to past consumption stimulus packages. The extension allows us to account for two novel features of the coronavirus crisis. First, during the lockdown, many types of spending are undesirable or impossible. Second, some of the jobs that disappear during the lockdown will not reappear when it is lifted. We estimate that, if the lockdown is short-lived, the combination of expanded unemployment insurance beneﬁts and stimulus payments should be suﬃcient to allow a swift recovery in consumer spending to its pre-crisis levels. If the lockdown lasts longer, an extension of enhanced unemployment beneﬁts will likely be necessary if consumption spending is to recover.”
Econ-Ark invites you to create your own model using these tools. The repository can be found here.
Julia Researchers Create Verifiable Neural Net in Julia for COVID-19 Epidemiology Model: In January, Christopher Rackauckas et al. published Using Differential Equations for Scientific Machine Learning, which introduced Universal Differential Equations (UDEs), combining the power of neural networks in Julia with the transparency of differential equations. In a new paper titled Quantifying the Effect of Quarantine Control in COVID-19 Infectious Spread Using Machine Learning, MIT researchers Raj Dandekar and George Barbarastathis describe how they use this method in Julia to create an epidemiology model of COVID-19 spread.
matplotlib is widely used to create scientific visualizations, one example relevant to COVID-19 is referenced in the paper “An analysis of SARS-CoV-2 viral load by patient age” Read the abstract below:
“Data on viral load, as estimated by real-time RT-PCR threshold cycle values from 3,712 COVID-19 patients were analyzed to examine the relationship between patient age and SARS-CoV-2 viral load. Analysis of variance of viral loads in patients of different age categories found no significant difference between any pair of age categories including children. In particular, these data indicate that viral loads in the very young do not differ significantly from those of adults. Based on these results, we have to caution against an unlimited re-opening of schools and kindergartens in the present situation. Children may be as infectious as adults.”
ObsPy is used, together with pandas and matplotlib in Jupyter Notebook, for processing seismic noise. ObsPy has been able to communicate about the effect of lockdown measures by using the recordings of seismometers worldwide. Fewer cars, less industry, and fewer planes equal fewer vibrations and less seismic noise. ObsPy has shared its code with seismologists so they can reproduce these results. Even seismologists who’ve never used Python before are able to use these tools effectively.
rOpenSci is referenced in various use cases around COVID-19:
- Noam Ross, Principal Scientist, Computational Research at EcoHealthAlliance is using the “drake” package to organize his processing of COVID data and for running simulations of hospital capacity for local government.
- Within the Imperial College London’s MRC Centre for Global Infectious Disease Analysis‘ response to COVID-19, Rich FitzJohn’s team are using the “cyphr” package for collaboratively working on sensitive data across a group of ~60 researchers, and the package “pdftools” for automated extraction of information from pdfs produced by public health agencies. Their work is primarily informing the UK government’s response, as well as other countries and NGOs.
- Nicholas Fraser and Bianca Kramer are extracting details from COVID-19 related pre-prints and visualizing their distribution over time. They are using rOpenSci packages rcrossref, rdatacite, aRxiv to facilitate this. Here is a citable version of the repository at figshare.
As the nonprofit umbrella organization for these projects, NumFOCUS provides crucial operational services on a daily basis to ensure that these open source tools remain accessible for future scientific discovery and innovation. Help us in our efforts to support our Projects by donating to NumFOCUS today.
Please contact us know if you are aware of any additional uses of NumFOCUS Projects helping to support scientific efforts involving COVID-19.