The coronavirus SARS-CoV-2 and COVID-19, the disease it causes, continue to be enigmatic. There are some who do not even realize that they are carrying the virus while others can not survive its attack. As India unlocks, the virus is reaching further pockets of the country with frailer healthcare systems. This makes it important to identify the potential sources of infection, and protect the vulnerable communities.

In the context of this article, we consider the most vulnerable communities to be those with the most aggressive symptoms, often needing hospitalization. On the other hand, those who carry the virus asymptomatically are the ones most difficult to detect, and are considered to be one of the primary sources of the virus. Had there been an affordable way to detect the presence of a virus by a personal device like those used to check blood sugar or blood pressure, things might have looked different for COVID-19 management. Unfortunately, in the absence of such technology, testing for coronavirus is done via multiple methods, each with different implications.

How do you detect a virus?

Looking for something as small as a virus is tricky. We can indirectly trace an infection by the symptoms it causes. But the symptoms of COVID-19 are varied, with symptoms like fever and cough that are not specific to SARS-CoV-2. So instead, we look for fragments of the virus in patient samples, which we can detect via molecular biology techniques. These fragments can be the viral proteins or genome - ribonucleic acid (RNA) - both of which can be found in nasal or throat swabs or sputum samples. Viral RNA can also be found in faecal samples of a small proportion of patients.

Our bodies react to infections by developing antibodies. The immune systems in our bodies detect pieces of the foreign (viral in this case) proteins, called antigens. Against these, our body makes a set of targeted proteins - antibodies - which are released into the bloodstream. This opens up the possibility of detecting viral infections through a blood test by looking for these antibodies or antigens.

RT-PCR - The gold standard method

The RNA sequence of the virus became available in February 2020. This allowed researchers to develop probes to look for specific sequences that could serve as a ‘signature’ for the novel coronavirus. In the Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) technique, the probes and enzymes amplify these sequences, which can be visualised and detected. These tests have been estimated to have at best 70% sensitivity. That means 30 out of every 100 positive samples can turn up as false negatives.

This test also requires facilities to have specialised PCR machines, trained personnel, and reaction kits that are largely imported – all adding to the cost of the process. Since the samples contain live viruses, testing is limited to facilities with Biosafety Level 2 or higher and is both labour- and time-intensive. All of these make RT-PCR a non-optimal choice for testing large numbers of people.

Antigen-antibody testing methods

Mass testing works best with techniques that require minimal infrastructure or training and are as fast and cheap as possible. Time-tested methods for these include testing for antigens or antibodies. Since antibodies bind to their respective antigens specifically, with a stock of antibodies, one can look for their antigen in blood samples, and vice versa. Both of these can be produced in large quantities, extracted and purified. Following purification, they can be immobilized on a plate or a paper strip for long-term storage and point-of-care usage. Antigen-antibody binding can be detected in multiple ways, and such tests have been deployed in diagnostics of various infectious diseases globally.

Though this may sound simple and straightforward, the reality of making a new antigen or antibody testing kit is different. To make an antigen-testing kit for SARS-CoV-2, one needs to select an antigen specifically expressed in this virus. To induce antibodies, it must be expressed in large enough quantities and should be visible to the immune system. These antibodies are generated in laboratory animals, extracted, and used for making antigen-detecting kits.

For an antibody-testing kit, we need to produce SARS-CoV-2-specific antigens using cells grown in the laboratory. This method allows large-scale production of the antigens but does not fully guarantee mimicking the native structure of these antigens. In such a case, it is difficult to predict whether they would bind with the antibodies produced in human bodies. Also, if the viral load is too low, the antigen test does not pick it up. A negative result in the antigen-test needs to be validated by RT-PCR – still the most reliable technique available.

Antibody-testing kits are known to be low on specificity. While designed to detect SARS-CoV-2, they can also detect antibodies produced against other coronaviruses. Antibodies against SARS-CoV-2 are elicited around two weeks after the symptoms appear in an infected person. So, if tested too early, one might not test positive though they carry the virus. An estimated 10% infected people do not test positive for antibodies. We don’t yet clearly know how long the antibodies against SARS-CoV-2, once formed, persist in the bodies. This means we don’t yet know how closely antibody-tests represent the present reality of the community’s health.

The Spike (S) and the Nucleocapsid (N) proteins of SARS-CoV-2 are the most preferred proteins to use as antigens in the antigen/antibody-testing kits. The N protein is the most abundant protein on this virus and thus induces antibodies easily. The S protein is the most specific protein to distinguish SARS-CoV-2 from other coronaviruses that can infect humans.

Different methods for mass surveillance

With economies opening up in the coming months, it is imperative to keep an eye on the community’s health through non-invasive, volunteer-independent surveillance techniques. This is especially important for a country like India with a large population earning its livelihoods through daily wages and pockets of high population density.

Mumbai, Pune, Delhi and many other cities have tested for antibodies via door-to-door collection of blood samples in select communities and extrapolated those results for the entire cities. These tests are increasingly being deployed in the country for large-scale testing. These results provide a basis for the local governments to evaluate the efficacy of their COVID-19 management strategies but depend on volunteer participation.

A non-invasive and volunteer-independent strategy involves checking for viral particles released into air or water by the infected individuals. Majority of this release is via respiratory droplets. But 30-60% of those infected also release non-infectious viral RNA through their faeces. While methods of air surveillance are still under development in India, sewage surveillance has already been done in cities like Chennai and Hyderabad. City sewage treatment plants record the quantity of sewage received and their geographic origins. The concentration of viral RNA in sewage water samples can be measured by RT-PCR. Through this, researchers can indirectly estimate the number of infected people in the city.

Pieces of SARS-CoV-2 RNA in faeces of infected persons can be found up to 35 days after the onset of symptoms. Thus, testing sewage containing faecal samples of a community, at any given time, paints a picture of the entire previous one month.

Why should we know this?

The results of these different testing methods are not always directly comparable. When Hyderabad’s sewage surveillance results were released, there was a lot of discussion about the unexpectedly large estimate of infected people in the city. However, the debates missed the intrinsic differences between direct and indirect testing methods that allow the latter to test much greater numbers of people.

Sewage surveillance can estimate infection levels in large communities at a negligible cost. However, it is largely limited to urban settings in India. On the other hand, healthcare workers are used to blood-based antibody tests, and the strategy can work in both urban and rural settings. These tests also give individual results, which can help in targeted demographic strategies for different genders, age groups and socio-economic backgrounds. But since these depend on people’s voluntary participation, they also tend to introduce bias in data. The study conducted in Mumbai, for example, had fewer women who got tested, which should be kept in mind by the policymakers who use the surveillance data to design their strategies in the coming months.

It is easy to be confused with all the different kinds of tests that are being done. The choice of tests is often made on the basis of tools and expertise available to a region. Their objectives can also span from individual to community management. Given the varying specificity and sensitivity of each of these tests, the estimated numbers of infected people in communities will differ in each method. With this article, we hope that if you or your community get tested, you will understand what is exactly being tested, its methodology, and interpret the results with clarity.

The COVID-19 pandemic drastically affected science and technology across the world. Conversely, it has also created many new scientific opportunities. Global scientific communities have been cooperating to further the progress of novel coronavirus (SARS-CoV-2) research and to identify new scientific activities to fill certain critical gaps. The need for continuous access to scientific data and learning is essential during this pandemic period.

Creative Commons and the Open COVID Pledge

Creative Commons (CC) is an organization that provides an open license to copy, distribute, and use intellectual property rights (IPR)-protected work. Recently, the CC organization implemented a call to promote open access in science to combat the COVID-19 pandemic. It requested scientists to follow zero embargo periods for their publications and adopt a CC BY and CC0 waiver for their research data.

CC0 means “no rights reserved”, which enables scientists to place their findings for free in the public domain. This allows others to freely enhance and reuse such works for any purposes without restriction under copyright or database law. CCBY enables researchers to remix, transform, synthesize derivatives, and rebuild the material for any purpose provided that they give appropriate credit to the original license holder and indicate what changes were made to the original material.

Restrictions in the usage of patents, copyrights, and other intellectual property rights (IPR) might end up costing lives during this pandemic. The Open COVID Pledge is an effort comprising scientists, advocates, entrepreneurs, and volunteers committed to resolving IPR-related obstacles in order to further the progress of research related to therapeutics and diagnostics for COVID-19.

The Open COVID Pledge calls for pharmaceutical industries, academic institutions and other organizations across the world to make their patents and Intellectual Property Rights (IPR) available free of charge for use in inventing new diagnostics, therapeutics, vaccines, equipment, and software solutions against COVID-19.

Further, the Open COVID Pledge requests individuals to raise awareness about the Pledge in their organizations and networks using the hashtag #OpenCovidPledge. Several big industries have already joined this pledge. IBM has shared their supercomputing power and AI for virus tracking and created a consortium to give researchers free access to over 400 petaflops of computing capacity. Furthermore, IBM has offered free access to its patent on touchscreens that use UV light for preventing pathogen transmission.

RADVAC, Mitsubishi Electric Research Laboratories, NASA-Jet Propulsion Laboratory at CalTech, Sandia National Laboratories, New Jersey Institute of Technology and many other laboratories and organizations have pledged their patents and copyrights under the Open COVID Pledge.

Open access for COVID-19 related publications

Several scientific publishing houses have been providing free access to scientific publications related to SARS-CoV-2 infection. Elsevier, a major scientific publisher, made its research and data related to COVID-19 freely available from March 13, 2020, at PubMed central and WHO COVID database. Further, Elsevier Connect created a COVID-19 Information Centre with the latest research information on SARS-CoV-2 and made more than 19,500 articles freely available via the ScienceDirect platform.

Moreover, Elsevier clinicians have also been curating data from Elsevier medical journals, textbooks, clinical information, as well as resources from other information providers and major health and government organizations for use by researchers, clinicians, and healthcare professionals. Similarly, other leading scientific publishers like the American Chemical Society and Springer Nature have also committed to supporting direct access to research and data available on their platform by the global scientific community.

Springer Nature has enabled free access to over 60,000 research articles, book chapters, and assay protocols on their platform. During this period, it has published about 10,000 new research/review articles on the COVID-19 pandemic and made all the underlying experimental data sets freely available for re-use.

The United Nations Educational, Scientific and Cultural Organization (UNESCO) has also partnered with several organizations to enable open science in the fight against COVID-19. The UNESDOC Digital Library, a part of UNESCO, recently released a set of policy guidelines for the development and promotion of open access.

The COVID-19 Universal REsource gateway (CURE) was recently established by the Indian Statistical Institute (ISI) in India, and Redalyc in Mexico. These organizations verify the relevance and accuracy of openly-licensed scientific data about the virus from different sources. The Stephen B. Thacker Center for Disease Control and Prevention (CDC) Library also maintains an up-to-date COVID-19 database. This CDC library allows researchers to search for and download research articles on COVID-19 from multiple databases.

The iSearch COVID-19 Portfolio is a comprehensive, expert-curated source for publications related to COVID-19 maintained by the NIH. WHO's COVID-19 research article database also allows searching through multiple databases and downloading articles. The Public Health Genomics and Precision Health Knowledge Base provides up to date genomics and precision health information on COVID-19.

Preprint publications

Preprints are full research papers that are shared publicly before they are peer-reviewed and accepted by a journal. Major preprint servers like bioRxiv and medRxiv have posted thousands of studies related to SARS-CoV-2 since the pandemic’s outbreak. Chemistry-related preprint servers like ArXiv and ChemRxiv have also shared several papers on COVID-19.

Preprints are initially examined by in-house staff and volunteer academics for plagiarism, biosecurity risk and completeness. After scrutiny, these preprints are published publicly within 2-3 days. The medRxiv publication duration is usually five days as it maintains a more in-depth screening process given that its publications are more directly relevant to human health.

However, there is also a global concern about non-peer-reviewed publications as they may influence public health policy decisions. For example, there have been instances of entirely computation-based analyses for COVID-19 treatments, publications with conspiracy theories, publications contradicting widely accepted public-health advice, as well as those using unprofessional language, eliciting severe concern about preprint servers.

Conclusion

There exist a multitude of scientific responses to the COVID-19 pandemic. Considering the cost of lives, several companies and investigators have waived off the usual IPR protections and restrictions through the Open COVID Pledge. Moreover, many scientific publishing companies have made COVID-19 related publications freely available online during this time. Furthermore, many researchers took advantage of the expedited framework for preprint publications during this period. However, additional scrutiny and scepticism are required while considering preprint servers as a resource for clinical applications.

Further Reading

• Bagdasarian, N., G. B. Cross, and D. Fisher. 2020. “Rapid Publications Risk the Integrity of Science in the Era of COVID-19.”BMC Medicine
• Glasziou, Paul P., Sharon Sanders, and Tammy Hoffmann. 2020. “Waste in Covid-19 Research.”The BMJ
  
• Kwon, Diana. 2020. “How Swamped Preprint Servers Are Blocking Bad Coronavirus Research.”Nature.
  
• McCreary, Erin K. and Jason M. Pogue. 2020. “COVID-19 Treatment: A Review of Early and Emerging Options.”OFID

“The human brain has one hundred billion neurons, each neuron connected to ten thousand other neurons. Sitting on your shoulders is the most complicated object in the known universe.”
~ Michio Kaku, theoretical physicist

Neurons, or nerve cells, have commanded a lot of attention in the scientific study of the nervous system. However, the nervous system also comprises another set of cells called “glia” that are just as abundant as neurons (if not more). Originally, glia were considered to function only as passive helpers to neurons. Now, however, these cells are pushing us to consider an overhaul of the perspective with which we view the architecture of the mind.

In recent years, scientists have begun trying to understand the crosstalk between neurons and glia and how they work together to develop the neural circuits in our brain. Although the study of neuron-glia interactions is still in its infancy, the picture is becoming clearer with time. The intricate nature of neuron-glia interrelationships can be seen in the connections between neuronal branches and those of astrocytes, a type of star-shaped glial cell. At these junctions, information is actively exchanged between neurons and glia.

The three basic properties of any information processing system are receiving a signal, processing it, and producing an output. We already know that neurons are the masters of all three. But can the same be said for glia?

A major question raised during the “glial revolution” was regarding the electrical properties of glia, or lack thereof. Neurons are regarded as the main information-processing cells in the brain because their membrane can rapidly and reliably conduct electric currents with the help of an array of ion channels expressed on their surfaces. These channels act like pathways that allow movement of ions in response to fluctuations in the electric potential difference across the membrane. These currents are, in turn, passed on to other neurons through junctions called synapses. Glial cells, on the other hand, express significantly lower levels of some of these key channels, suggesting that they lack this important property of electrical conductivity.

The past few decades have witnessed the advent of several new molecular biology techniques. Armed with an arsenal of modern tools, researchers are now able to directly record the electrical activity of individual cells. In the late 1980s and early 90s, researchers using calcium imaging observed that fluctuations in the levels of calcium ions in glial cells alter their electrical properties. This revolutionised our view of these cells. Changes in calcium ion levels along with the ability to chemically communicate with neurons make glia an undeniable component of neural circuits.

Computer scientists, too, were not oblivious to this increasing evidence supporting the importance of glia. Artificial Neural Networks (ANNs) have been in the picture since the 1950s. These are information processing systems that try to replicate the way our brain thinks by simulating biological networks. The earliest of these models were capable of producing a single output signal by analysing the inputs received, much like their living counterparts.

This paved the way for computational networks which could be “trained” to learn the relationship between an input and its output. Artificial Neural Networks today are composed of complex layers and webs of computational nodes, also known as “neurons”. These are connected in intricate ways and can process enormous quantities of data to pick out similarities or patterns, a feat previously thought to be unattainable by a machine.

Owing to the emerging roles of glial cells, efforts have been made to extend these ANNs to incorporate glial cells as well. Where does India stand in terms of research in this exciting field? To understand this, I had a conversation with V Srinivasa Chakravarthy, Professor at Indian Institute of Technology (IIT), Madras, whose illustrious career has spanned over two decades. He entered this field as early as 2010 when his group proposed an “energy-matching” principle while studying blood flow to the brain.

Energy is delivered to neurons from blood vessels in the form of lactate or lactic acid passing through astrocytes on the way. During this process, the total energy that moves into the neuron should match the neuron’s energy demand. Otherwise, the normal functioning of the neuron could be disrupted. The models developed by Chakravarthy’s team suggest that there must be some kind of learning going on in these networks, both at the astrocyte-blood vessel and the neuron-astrocyte levels. It was the first time anyone proposed Hebbian learning (explained aptly by the adage “neurons that fire together wire together” ) and plasticity (the ability to change dynamically) in such networks.

The focus on artificial neuron-glia models in India does not seem to be very widespread. According to Chakravarthy, research into artificial neural networks essentially started with simple computational models and took a few decades to move towards more complex models. However, glia were modelled in detailed biophysical terms almost from the very beginning, which may have resulted in a slow pace of research in this area, a situation akin to missing the forest for the trees.

Essentially, since glial cells had been overlooked for a long time, the biological principles that govern their function were not completely clear. However, a somewhat promising realization from neurobiology was that astrocytes, like neurons, can regulate processing and transfer of information.

In an interdisciplinary study undertaken by a group of researchers at the Universidade da Coruña, Spain, in 2012, single “astrocytes” were added as modulatory units to individual neurons in an artificial neural network. These computational glia could change the relative weights of the input or output signals or both. This system was then applied to complex problems such as disease diagnosis. This network inherently relies on the properties of astrocytes and research in this direction is currently ongoing.

What are the implications of these models for Artificial Intelligence (AI)?

Looking at it from a biological angle, Chakravarthy explains, “I like to think of them [glia] as conduits between blood vessels and neurons. Vessels are carriers of huge amounts of energy, but neurons don’t have the storage to support this quantity. So, several buffers would be required at each stage.”

In the context of AI, networks can be implemented both in software and hardware. Clever computer programs need a physical shell to be fruitful, such as microchips. Hardware implementation in these cases relies on numerous parallel computations to accelerate the performance of the software. This, in turn, makes energy conservation crucial to prevent the devices overheating. Normally, in a chip, a power line fuels all the circuits. But in a futuristic realization, we could perhaps have a branching pattern of the power line, similar to blood vessels in the brain. To feed current from these “power trees” into local neuron circuits, we would need an intermediary to temporarily buffer and redistribute the energy properly. Digital equivalents of the glial cells could serve as these buffers.

“The only job is to look at them [glia] in simpler ways without getting lost in the microscopic details,” advises Chakravarthy. Obtaining a complete understanding of neuronal networks is an inconceivable feat. Neuron-glia interactions are even more puzzling given the paucity of available data and the increased complexity of the system. The role of glia in higher-order brain functions such as cognition also remain incompletely understood.

We can hope that as biological research in neuroscience advances, the computational evidence to illustrate the relevance of neuron-glia interactions will synchronously gain ground. Perhaps, with advances in AI technology, we might even be able to successfully emulate the plethora of interactions in the brain someday.

If previous pandemics have offered the world a lesson or two, then it is the use of masks. The practice of wearing face-masks to prevent disease can be dated back to the 1910-11 Manchurian epidemic in China. It was Wu Lien-teh, a Cambridge educated Chinese doctor who described the mask as a ‘prophylactic apparatus’ that could be worn by all to protect themselves from the plague. An entire century later, masks have remained a humble yet effective tool that can be easily mass-produced and utilized in our battle against infectious disease.

In the same decade as the Manchurian epidemic, over 40 million people around the world lost their lives to the Spanish flu of 1918. This number is greater than the total casualties from World War I. The same outbreak, when it reached India, arrived as the Bombay fever and took 17 million lives, 40% of the total deaths worldwide. The practice of covering one’s face with scarves and veils with the intention to ward off disease originated during this period and remained until it faded towards the end of 1919.

In 1923, the Great Kanto earthquake in Japan was followed by a firestorm that wrecked over half a million homes. Damage to the Fukushima reactors leaked radiation and left the sky filled with residual ash for months. Facemask production surged and it became an accessory for the inhabitants of Tokyo and Yokohama. The influenza epidemic in 1934 triggered a different relationship with masks for the Japanese, that of social courtesy. Infected people were conscientious about not passing on germs to others, and masks were no longer only worn by healthy individuals trying to avoid illness.

This early, occasional mask-wearing transformed into a year-long habit as industrialisation from World War II and pollen emitted by the Japanese cedar polluted the air. Today, this practice is deep-rooted in the historical framework of East Asian culture. It symbolises more than just protection from infection and represents cooperation and solidarity that allows a society to function during a pandemic. It reinforces civic duty and helps us deal with the fear that conquers the healthy and sick alike.

Mask wearing is not new, COVID-19 is

Today, the most affordable protection available against the novel coronavirus which causes COVID-19, is in front of our faces, literally. There is an extremely good reason to wear a mask: wearing one decreases the probability and the number of people who can be infected by a person who carries the disease. Masks are designed to prevent the escape of a significant number of droplets (also known as aerosols) expelled from our nose and mouth. A cough can produce as many as 3000 droplets and I may spray thousands of microscopic droplets as I utter the phrase - “Wear a mask!”

The N95 mask, for instance, can block particles that are as small as 0.3 microns with 95% efficiency, while surgical masks can do the same with an efficiency of 75-80%. Although not all masks offer complete protection against the entry of viral particles, they still aid the immune system by effectively reducing the overall quantity of viruses that the immune system might have been exposed to otherwise.

This becomes relevant when it comes to barricading virus-containing droplet nuclei using masks. Droplet nuclei consist of tiny amounts of fluid, carrying the virus that remain suspended in the air for several hours. They are usually less than 5 µm in diameter – twenty times smaller than the width of a human hair. As a result, these droplets can travel by virtue of air currents and can be breathed in, as opposed to larger droplets that usually fall to the ground or a surface due to gravity. Thus, these small droplet nuclei can act as a vehicle for airborne diseases. Droplet nuclei can be formed by the natural drying of respiratory droplets that are generated naturally as a result of talking, breathing, coughing etc.

The principal route for the transmission of the novel coronavirus is through respiratory droplets from pre-symptomatic as well as asymptomatic individuals. Although the viral droplets were earlier thought to remain actively suspended in air for only 3 hours, a new preprint suggests that the viruses in the droplets can survive for over 16 hours.

Some of the measures that can prevent the spread of the virus are limiting interpersonal contact through physical distancing and tracing infected individuals. Wearing masks in public has been established as effective in reducing the incidence of transmission per contact. As one would expect, this is most effective when mask-wearing is adopted as a consistent practice by the public.

Respiratory protection for respiratory transmission

A research team led by Jeremy Howard, data scientist and faculty member at the University of San Francisco, proposed a mathematical model of COVID-19 transmission. The model estimates the successful outcome of wearing masks considering three aspects: efficacy of the mask in offering protection against the virus, the proportion of the people who practise mask-wearing, and the rate of transmission of the virus. The model suggests that if most people wear a mask in public, the transmission rate is at its lowest, effectively slowing down and eventually stopping the spread of the disease. A rudimentary DIY (do-it-yourself) mask will not block every single viral particle but it lowers the transmission rate.

As asymptomatic, infectious carriers contribute to a significant proportion of the infections and masks do an effective job of blocking the infection at the source, “masks for all” may be the best option for preventing disease spread.

In a 2015 study, a team of Indian scientists studied the dispersion of particles expelled during coughing/sneezing using aerodynamic simulation techniques. The study was led by Guruswamy Kumaraswamy, Indian Institute of Technology (IIT) Bombay; Prem Andrade, Ansys Software; and Pankaj Doshi, Pfizer Inc. The researchers observed that larger droplets emitted from coughing/sneezing can be blocked by a mask. A mask reduced the distance travelled by smaller droplets from nearly 2 metres to less than 30 cm. At the same time, maintaining a physical distance of at least 2 metres is greatly beneficial as even the particles that escaped the mask in the study could carry the virus no farther than 1.5 metres.

A Cambridge University study published in 2013 compared homemade masks made out of a variety of household materials with surgical masks on their efficacy in offering protection during an influenza pandemic. All the masks studied in the report reduced the number of microorganisms expelled into the air by volunteers, at least to an extent. Masks from dish/cleaning towels or cotton blend t-shirts turned out to be considerably effective in capturing small particles (stopping 83% and 74% of the particles, respectively).

Cloth masks, when worn by infected individuals, can play a role in protecting people in their immediate surroundings from the infection even if they are not as effective as an N95 respirator tailored for medical personnel. Cloth masks offer source control - they reduce the possibility of transmission of infection from an infectious mask wearer. This is the single most important reason for the public to wear masks.

In order to rightly prioritise the medical and clinical community’s access to respirators and surgical masks, the Indian government is encouraging the use of cloth masks. Taiwan, Thailand and the Czech Republic are other governments who have encouraged the use of DIY or cloth masks by the public. Cloth masks, as an alternative to surgical or N95 masks, offer reduced blockage of viral particles and yet a higher degree of protection against transmission when a large proportion of the public wear them.

Mask-wearing is an enduring and powerful form of protection, even beyond the biological context of disease spread, as it fosters a sense of social responsibility and unity, which is necessary to face a challenge like the present pandemic.

On a hot, sultry, and a rather quiet afternoon amidst the COVID-19 scare, I am on my way with Dhiviya Vedagiri to the Biosafety Level-3 (BSL-3) lab at CSIR-Centre for Cellular and Molecular Biology (CCMB) where she and others have been growing SARS-CoV-2, the coronavirus responsible for the COVID-19 pandemic.

Dhiviya, a PhD student with Krishnan H Harshan, works on understanding the interplay between flaviviruses and host immune responses. However, that is now on a temporary halt in order to channelize all efforts into the fight against COVID-19. Growing the virus is the first step towards developing vaccines against the virus as well as testing potential drugs.

The BSL-3 is almost a sacred place – dedicated to culturing disease-causing infectious microbes. While growing non-pathogenic E. coli bacterial strains is a fairly commonplace activity in many life science labs, infectious ones need a plethora of safety rituals to be followed.

The journey to a BSL-3 lab begins with a series of change rooms. Across these, the air pressure keeps dropping, and the amount of Personal Protective Equipment (PPE) on the entrant keeps increasing. A consistently lower air pressure than the outside change rooms (by 30Pa – not something that you would generally feel) ensures the potentially contaminated air from inside does not go into the change rooms. But the seriousness of the situation dawns upon you with the PPE – foot covers, special shoes, shoe covers, a lab coat followed by a surgical gown, gloves, a pair of goggles, an N95 mask and a head cover – all to ensure that no body part is exposed to even this “clean” air. The focus is on making sure that the researcher neither carries in any contamination from outside nor brings out an infection from the BSL-3 while leaving.

Entering the BSL-3 finally, we are greeted by bright light and shiny clean stainless steel furniture housed in a temperature and humidity-controlled lab. We don another pair of gloves and move towards the incubator where the SARS-CoV-2 virus particles are growing on their host cells in little wells in a plate placed in a temperature-controlled incubator. The viruses are kept cosy at 37ᵒC with controlled amounts of CO2, their host cells sustained by a few millilitres of a tomato red coloured nutrient solution, supplemented with proteins and finished off with a dash of antibiotics to keep bacteria at bay.

A virus is a quasi-alive particle carrying minimal content with itself. Its genetic material is encapsulated within lipid and protein layers. It does not grow in size or metabolize nutrients, but it multiplies in number within a host whose protein machinery it can hijack for its own use. SARS-CoV-2 can infect cells that express a certain set of proteins called ACE-2. Here in the BSL-3, they are offered ACE-2 expressing kidney epithelial cells from the Green African monkey (sold as Vero cells) to grow on.

Dhiviya comes here every few days to see how the Vero cells infected with SARS-CoV-2 are faring. A change in the colour of the nutrient solution indicates a change in the pH (acidity) of the solution. She would then need to monitor the CO2 settings in the incubator or add fresh medium. Today she takes out a 6-well plate containing the Vero cells she had infected with a sample isolated from a COVID-19 patient a week back. The colour of the nutrient solution appearing fine, she places the plate under a microscope to check the health of the cells.

The Vero cells grow flat on the plate’s surface and appear irregular in shape. But most of the cells in the sample she is checking have lost contact with the surface and can be seen floating as globules. “These are dead cells,” she says. The viruses from the patient sample seem to have infected the Vero cells, killing them. It was now time to confirm it with RT-PCR.

RT-PCR allows researchers to check for the presence of the virus’ genetic material – its RNA - in a sample. RNA is made of four repeating subunits called nucleotides– adenine (A), uracil (U), cytosine ( C) and guanine (G). However, SARS-CoV-2’s RNA sequences are far longer than most other RNA viruses with about 30,000 A, U, G and Cs. It is this sequence that is read by the host cell’s protein synthesis factories and used to make viral proteins. These new proteins then make many copies of the virus’s entire RNA to be packed inside lipid and protein membranes, which are released as new virus particles. The sequence of this RNA can mutate as the virus grows, and thus, alter its proteins – a part of its constant evolutionary arms race with its host. This change can sometimes alter the virus’s infectivity or the severity of disease symptoms it causes.

An effective vaccine against SARS-CoV-2 would be one that targets most, if not all, COVID-19 causing strains. An effective drug testing platform would need as many of these strains to be grown and tested as possible. Both of these require millions of copies of the virus to be grown, collected from at least hundreds of patients to cover as many strain varieties as possible. It is an activity that is expected to take 2-3 months in the very least at CCMB to have a fair representation of the Indian coronavirus.

Dhiviya takes the 6 well-plate to a working hood which has a continuous exhaust with HEPA filters that trap the viruses before the air goes out. Times moves slowly here. Every step in the procedure begins with cleaning surfaces and equipment with bleach, alcohol, and UV radiation. A few hundreds of microlitres of the nutrient solution from the wells of infected and uninfected samples are collected into tubes. The virus has two protective layers – an outer lipid bilayer, and an inner protein capsid to which its RNA is bound. RT-PCR looks for this RNA. So, the protective layers have to be broken open.

Alcohol and detergents dissolve the lipid membrane and puncture the protein layers. Once the membranes are broken, the cell soup is gently transferred onto a bed that specifically binds the RNA. Dhiviya washes the bed to remove remnants of salt and alcohol as these might cause problems during RT-PCR. She then adds water to the bed to dissolve the RNA and collects it for running the RT-PCR reaction. The foremost concern here in what is otherwise a regular molecular biology lab practice is to avoid any risk of aerosol formation or spillage.

Now that the virus is broken down and hence inactive, the sample doesn’t need to remain under BSL-3 conditions. Dhiviya puts the RNA sample dissolved in water in an enclosed chamber between BSL-3 and the more relaxed BSL-2 labs – called the “passbox”. Samples can be moved out of BSL-3 only after they go through a final UV irradiation within the passbox itself. Then, with a final clean-up of her working surface and her equipment, complete with a dose of UV, Dhiviya and I finally move out of the BSL-3.

Life without the PPE feels so much lighter and pleasant. The gowns, headcovers and goggles go for recycling - they would be treated with high temperature and pressure to kill any living organism on them (a process called ‘autoclaving’). The lab coats go for UV treatment, while the N95 masks, gloves, shoe and foot cover go into a discard bag. Any discard from the lab or the change area will be autoclaved before disposal.

I head for home, and Dhiviya catches some fresh air. Later in the evening, she runs the RT-PCR and checks if the samples show the same sequence of A, U, G and C that SARS-CoV-2’s RNA contains. As I wrap up my dinner, Dhiviya’s message flashes on my phone – “The samples are positive.” One batch of the virus is now growing in the Vero cells, with many others in waiting to be eventually grown in quantities useful for vaccine development and drug testing.

A novel acute respiratory infection changed the global perspective of healthcare in December 2019. Humanity is currently facing the challenge of the century - the COVID-19 pandemic. The World Health Organization (WHO) has declared an emergency and governments around the world are devoting a significant amount of attention towards controlling the spread of this disease. The most commonly adopted strategy to contain the disease is a "lockdown" aimed at preventing a chain reaction which can wreak havoc on the medical infrastructure of any country.

In addition to governments, scientists and doctors are at the forefront of the battle against the novel coronavirus. Many researchers and medical professionals are toiling constantly to find a solution to this potentially fatal disease. This article focuses on some recent updates towards the development of new therapeutic strategies and vaccines against COVID-19.

The Novel Coronavirus

Viruses are chemical entities (non-living infectious agents) that need a cell to multiply and survive. SARS-CoV-2, the virus which causes COVID-19 has single-stranded RNA as its genetic material. Its genome is very small, containing only 15 coding genes and ~30000 nucleotides. In contrast, humans have approximately 30,000 genes and over 3 billion nucleotides in their genome.

Current therapeutic strategies

Currently, the most common treatment strategy is to relieve the patients’ symptoms (which resemble pneumonia), while the hunt remains on to find a complete cure. One approach to devising therapeutic strategies is to unveil the biology of the virus - its structure, how it causes disease (pathogenesis), how it infects people, and how the disease progresses.

SARS-CoV-2 attacks vital organs such as lungs, heart, intestine and blood vessels. In the lungs, the virus targets cells present in the lining of the lungs (called pneumocytes) and this results in respiratory distress. This, in turn, leads to a decrease in oxygen levels in the blood and finally to death.

A recent preprint report reveals that the virus interferes with heme, an iron-containing compound which is an important component of blood. Another study, also in the preprint stage currently, suggests greater susceptibility of patients with blood group A to SARS-CoV-2 compared to other blood groups. Additionally, there exists evidence that patients with heart disease and diabetes are more susceptible to this disease. The above study also suggested that certain treatments for both these disorders lead to over-expression of a protein called angiotensin-converting enzyme 2 (ACE2), which SARS-CoV-2 can bind to and use to access host cells. Hence, this further exacerbates the infection risk for these patients.

A pair of recent studies published in Science provides a clue towards understanding what makes SARS-CoV-2 so infective. Like all coronaviruses, SARS-CoV-2 has a spike-like structure on its surface, which gives it an appearance like a crown. These studies suggested that this spike-like structure binds tightly to human cells - much more tightly than the coronavirus that caused the Severe Acute Respiratory Syndrome (SARS) outbreak in 2003.

Apart from understanding the biology of the virus, some drastic steps are needed to restrict the contagion in the current scenario. SARS-CoV-2, being an RNA virus, can be inhibited by drugs previously used for other RNA viruses, e.g. the Human Immunodeficiency Virus (HIV). Trials are currently underway with a combination of two anti-HIV drugs - lopinavir and ritonavir.

Similarly, the National Institute of Infectious Disease (Japan) demonstrated the use of a drug that was previously developed against the SARS (2003) coronavirus. Additionally, a combination of drugs including chloroquine, a potent medicine used to treat malaria, has also been proposed for clinical usage. It is hypothesized that this combination can prevent the virus from binding to heme. Also in the pipeline are several drugs that are in various phases of clinical trials (Table 1). These include approved compounds such as Kevzara, a rheumatoid arthritis drug that decreases lung complications, successfully tested in COVID-19 patients.

A drug, starting from the time of its inception and research to qualifying all three phases of clinical trials, takes almost 10 to 15 long years to come to the market. Nonetheless, there is a possibility that coordinated international efforts and availability of good funding may make drugs against COVID-19 available within a record time. There is also hope for development of new therapies such as monoclonal antibodies, which may take less time to be available to the doctors due to their speedy trials and their high specificity.

Vaccine Development

Vaccines stimulate the immune system of an individual to prepare it against a future pathogen attack. This is done by pre-exposing the person to either killed or weakened pathogens, or some of the pathogen’s structural parts, which leads the body to mount a defence response.

For the development of a vaccine against SARS-CoV-2, a similar approach is under consideration by Serum Institute of India and Sanofi Pasteur, France. An alternative strategy is also to generate antibodies against the spike proteins of the virus, which is being followed by Moderna Inc., MA, USA. Additionally, a German Enterprise, CureVac, aims to design an RNA-based vaccine against the virus. In this approach, RNA that codes for some of the viral proteins is introduced into the body. This RNA can be used to produce viral proteins, against which the body can then synthesize antibodies, thus preparing for the virus’s attack.

All these studies are under different phases of clinical trials (Table 1). These vaccines may become available in the near future, but the time it will take for these to reach the market depends on the efficacy and success in all three phases of clinical trials.

Conclusion

The COVID-19 pandemic has caused an enormous amount of financial and social burden. Research has accelerated, but is still in infancy and will require time and funds to translate into therapies and vaccines. Pandemics like the current outbreak disrupt developing countries that have inadequate financial capabilities or a fragile healthcare system. However, there is hope in the fact that countries across the world have been uniting to fight this challenge. As a citizen, our obligation is to follow official advisories, not believe in myths, and educate ourselves.

Table 1: Some companies currently developing drugs/vaccines against COVID-19

Data Derived from: “An updated guide to the coronavirus drugs and vaccines in development” by Damian Garde, STAT News, March 2020

In the wake of the recent global health threat from the novel coronavirus, the World Health Organization (WHO) issued immediate counter-measures to control the spread of the disease. Have you wondered how experts gain insights that lead to timely decisions? Or how they arrive at estimates of the number of people who might get infected? How do experts visualise the progress of the infection and its capacity to cross borders? Based on what conclusions, are lockdown directives issued? And finally, how can experts predict how long an epidemic will last?

The answer lies in maths. Mathematical functions can be applied as tools to describe the dynamics of how infectious diseases propagate among people. Mathematical Modelling generates a picture or a ‘model’ of the dynamics of the disease, which can be visually represented by graphs, charts and comparative tables.

Models provide valuable inputs to visualise how diseases affect people. Hence, epidemiologists — public health experts – use them extensively to assess risk or to analyse intervention strategies to control or prevent diseases. Insights available from models facilitate disease management protocols like mass vaccination drives, treatment patterns, and precautionary procedures.

When the infectious disease is an unknown one, such as the present coronavirus pandemic, models become all the more vital for policymaking. “Models can help answer several questions that impact policy. In most cases, they are the only rational way of formulating such questions and evaluating how different interventions might shape the spread of the disease,” says Gautam I Menon, Professor, Departments of Physics and Biology, Ashoka University (Sonipat) and Institute of Mathematical Sciences (Chennai).

The hidden patterns

Mathematical models have century-old roots. In the 1920s, William O Kermack and AG McKendrick observed that a population that is exposed to an infection can be divided into three categories– Susceptible, Infected, and Recovered. They found a way of representing the numbers in each of these groups mathematically.

They translated their idea into differential equations, which draw a relationship between a physical quantity and its rate of change. The Kermack-McKendrick equation estimates what fraction of the population, over time, enters into one or the other of these categories, starting from an initial state in which one infectious person ‘seeds’ the infection among the rest.

From this, Kermack and McKendrick devised their classic SIR (Susceptible-Infected-Recovered) model that could predict disease spread. Since then, mathematical models have played a prominent role in transforming public health care. Governments, health organisations, scientists and hospitals depend heavily on models to deal with the onslaught of issues that arise out of medical problems.

“Virtually all other models in use today build on the intuition of the SIR model but often introduce additional categories,” says Menon.

Compartments, Networks and Agents

Several factors govern the transmissibility of the infection from the affected to the unaffected. The spread can be through direct contact or through water, air, or surfaces which harbour the pathogen. Also, disease dynamics can be studied at different scales: the single individual, small groups of people, and among entire populations. Different models are chosen based on the complexity of available data. In their modern avatar, models are simulated by computers that generate the numbers and distribution patterns of infections.

In the simple SIR model, people fall under any of the three ‘compartments’ – Susceptible, Infected, or Recovered. The equations describing them assume that the Infected can interact with the Susceptible, infecting them and converting them into Infected as well. As the Infected increase, the Susceptible decline. Infected people can also recover, and are then assigned to the Recovered compartment.

These equations can then be solved to understand how the number of infected people changes with time. Sometimes, additional categories are introduced to the basic model. For example, an infected person who is not displaying any symptoms can form a new category. Sometimes, additional factors such as age also need to be taken into consideration.

But reality presents many additional degrees of complexity. People mix randomly and socialise, which has a marked effect on disease propagation. Crucial aspects such as infection progression, the degree of susceptibility of the person, or demographic factors are left out in the basic model, calling for more comprehensive models.

Some diseases (such as measles or sexually transmitted diseases) spread quickly due to social or physical interaction among people. A network model takes these additional aspects into consideration. Here, the individuals are the ‘nodes’ while their interactions form the ‘links’ of a network. “For example, members of close-knit families, teachers and their students, or doctors and their patients are linked through some physical proximity that can propagate the disease, forming well-networked people with many links,” explains Menon.

People mix randomly across geographies as well, resulting in the disease crossing borders. Demographic differences, social influences, and predispositions also influence disease propagation. Some of this diversity can be simulated using a computer program and checking for the probable outcomes when the many variables interact with each other.

Such variables can be incorporated into an agent-based model that allows researchers to analyse patterns of epidemic spread in these more complex situations, which can then be compared to the observed data. By incorporating factors that describe how an individual responds to interaction, environment, place, and time, the model estimates the probable spatial distribution of the disease and its evolution in time.

In the agent-based model, each individual is an ‘agent’ defined by particular attributes. For example, an agent might be at different geographical locations, e.g. work or home. Then a computer program is used to simulate the interactions of the agents and the spread of disease among agents. This model offers flexibility in deciding what goes into the definition of an agent and the way agents transfer diseases among themselves. “For example, each agent could have a different immune response to the disease, or the age of the agent could be relevant,” elaborates Menon. He also points out that such intricate modelling needs high-power computers to perform the number crunching.

Impacting policies

Such systematic processes for estimation of variables and parameters inherent to an epidemic proved invaluable in the case of the present coronavirus outbreak. Several models have emerged, uncovering the transmissibility and the destructive patterns of the pathogen. “One model used airline traffic data to find out how the infection might have spread to other regions outside Wuhan (the epicentre of the disease in China),” says Menon.

Such models can calculate the expected number of cases and how quickly the disease might spread, along with helping quantify quarantining and social distancing measures. They aid health services in budgeting medical equipment and facilities like ICU beds and ventilators, facilitating immediate policymaking.

“Although a comprehensive model specific to India is yet to emerge, the Indian government seems to be following most of the WHO guidelines carefully and constantly reassessing its response,” says Menon. He emphasises that social distancing is one way of controlling disease spread by changing the way people interact with each other. By exercising a collective responsibility in implementing the measures, we can avoid overwhelming the medical services or clogging them, he concludes.

In recent times, we have seen an increasing number of instances of Artificial Intelligence (AI) donning the proverbial lab coat. In early 2019, thousands of people were screened every day in a hospital in Madurai by an AI system developed by Google that helps diagnose diabetic retinopathy, a condition that can lead to blindness. Startups like Niramai, based in Bengaluru are developing AI technology for early diagnosis of conditions like breast cancer and river blindness.

The sudden, accelerated growth of Machine Learning not just in research but in all walks of life can bring to mind Black Mirror-esque visions of dystopia in which machines rule over humanity. But let us leave worrying about the consequences of the far future to science fiction and look at the immediate impact this technology has had in science.

The terms Artificial Intelligence and Machine Learning are often used interchangeably, but there is a slight distinction. AI can be used to describe any machine capable of performing tasks we consider requiring intelligence, such as playing chess.

Back in 1997, IBM’s Deep Blue computer forced its way through over 200 million computations per move to defeat the reigning chess world champion, Gary Kasparov. This event created similar wild predictions of global machine dominance. However, Deep Blue was an early supercomputer with less processing power than the smartphone in your pocket today. While Deep Blue was completely programmed by humans, it could very well be called an artificial intelligence. However, it involved no aspect of Machine Learning.

Machine Learning provides computers with the ability to develop ways of solving problems without being programmed to do so. Machine learning algorithms are designed to identify patterns and make decisions the way humans do. This simple shift in programming from instruction to learning has led to some remarkable results. However, the question worth investigating is, why now? Why has this particular decade led to the evolution and proliferation of such powerful algorithms?

In order to find an answer, I spoke with Sudhakaran Prabakaran, a lecturer at the Department of Genetics at the University of Cambridge, as well as an Assistant Professor at the Indian Institute of Science Education and Research (IISER), Pune. My first hunch was that the growth of AI in research today could perhaps be attributed to recent developments in their algorithms. Consider the following excerpt written in Spanish, that can be expertly translated by a machine:

Hoy en día los algoritmos de inteligencia artificial son extremadamente sofisticados y son utilizados en todas partes. Probablemente está leyendo una traducción en inglés de este párrafo dado por el traductor de Google. En el 2016, el traductor de Google cambió todo su código a un sistema de traducción automática neuronal - un cambio que ha reducido el tamaño de su interfaz de cienes de miles de líneas de código a solo 500. Sin duda este cambio ha dado mejores resultados comparado con las frases extrañas dadas antes.

Translation by Google translate:

Today, artificial intelligence algorithms are extremely sophisticated and are used everywhere. You are probably reading an English translation of this paragraph given by the Google translator. In 2016, the Google translator changed all its code to a neural machine translation system - a change that has reduced the size of its interface from hundreds of thousands of lines of code to just 500. No doubt this change has given better results compared to the strange phrases given before.

Despite their present-day success, it turns out that the inception of machine learning algorithms goes far, far back in the past. Prabakaran notes that the very first machine learning algorithms were designed as early as the 1950s. These were called neural networks since they mimicked the way neurons are interconnected in our brain.

A neural network is given a set of information as input such as a phrase in Spanish and is expected to predict an outcome like an English translation of the phrase. ‘Learning’ proceeds by a process similar to trial-and-error, except that each try is influenced by feedback from the previous tries. The network tests its performance against the expected set of results and retries the prediction with slightly modified internal settings. This process is repeated until the network finds the seemingly magical configuration that gives the right answer almost every single time.

Today’s machine learning algorithms come in many more different avatars compared to basic neural networks, but their underlying principles remain the same. These algorithms have very interesting names, like convolutional neural networks, genetic programming, random forests, lasso and so on. Yet the frameworks for even these advanced algorithms were already established well before the turn of the millennium.

Why then was it uncommon to see them being implemented in scientific research? Did we lack relevant research questions to apply these techniques to?

Not quite.

Prabakaran gives the example of a very early problem-solving technique that was used in London way back in 1854. The city was facing a deadly cholera epidemic, the cause of which was unknown at that time. In an effort to control it, several scientists were studying the outbreak closely to figure out where it started and how it spread.

The breakthrough was made by a physician named John Snow, who drew a detailed map of the Soho district in London showing places where individual cases of cholera were recorded. Far from knowing nothing, this John Snow soon arrived at the big picture by meticulously arranging all available information. He realized that most cholera cases were clustered around certain water pumps. This led him to correctly correlate the disease to contaminated water supplied by these particular pumps.

What John Snow did is exactly what AI is so good at - clustering information and finding patterns. There have always been research questions which required effective use of these techniques. What John Snow and the early pioneers of AI did not have access to, was fast and efficient computers. Is that what lead to the AI revolution?

Yes, but that is only part of the answer.

The advent of the silicon age in the 80s and the 90s made faster, cheaper and smaller computers available to everyone. The steep growth in processing power helped create leaps of progress with the use of statistical data analysis techniques and automation in scientific research. Yet, for a long time, the AI algorithms remained relatively in the shadow.

To complete the answer, notice that we have been gradually uncovering more and more information in order to find out what makes machine learning effective today. This is exactly what makes machine learning algorithms flourish! We need data, and in immense quantities.

“The datasets needed to train and test machine learning algorithms today are not produced by a single individual, lab or institution, but by entire consortiums,” explains Prabakaran. His group is involved with The Cancer Genome Atlas (TCGA) consortium, which has accumulated over 5 Petabytes of data about many different kinds of cancer - an amount of data so humongous that if we were to write it all on CDs and stack them vertically, it would reach the height of Mt. Everest!

According to Prabakaran, the biological sciences witnessed several inflection points when it came to gathering quantitative data. An important one came with the completion of the Human Genome project at the turn of the millennium. Starting in 1990, it took 13 years to complete sequencing the first human genome. 15 years hence, the number of genomes sequenced is now approaching 1 million.

It’s not just limited to genomic sequences. Consortiums worldwide are collecting data at a rapid rate and on stupendous scales. This includes structural information about proteins and other cell components, microscopic images of cells and tissues, MRI scans, satellite images of vegetation and oceans to name a few. It is estimated that the total size of biological datasets is now of the order of exabytes, which would make the height of our stack of CDs reach a satellite’s orbit.

Beyond refinements in the techniques involved in collecting data, we can also store and manoeuvre it like never before. Cloud computing is proving to be the key to enable numerous researchers across the globe access to analyse a common pool of data. This is instrumental in forging collaboration between different labs and institutions. Increased collaboration, especially with industry, has helped bring down costs. While the Human Genome Project cost a total of almost $3 billion, you can walk into a clinic today and get a scan of your genomic data for less than $1000.

Finally, it is the availability of open-source software for machine learning with a relatively gentle learning curve that really makes it possible for any individual to use them to draw some fantastic inferences from their data. “Even with relatively less computing experience, one can learn machine-learning modules based in programming languages like R and Python within months,'' says Prabakaran.

All of this has made it easy for researchers from a diverse pool of backgrounds - from experimental biologists to physicists and mathematicians to wear the hat of a data scientist using machine learning for their work. Examples include a November 2018 study from the Karnataka Forest Department in which AI helped perform a comprehensive survey of leopards in South India. In July 2019, researchers from the United States in collaboration with the Centre for Wildlife Studies, Thrissur, Kerala used Machine Learning to identify bat species which could potentially cause an outbreak of the Nipah virus.

Even if we just look at the last two years, the list of AI applications that have changed our perspectives towards the stories our gathered data is telling us is long. It is indeed exciting to think what the next decade will bring.

Sudhakaran Prabakaran and his colleagues Pranay Goel from IISER Pune and Leelavati Narlikar from the National Chemical Laboratory (NCL), Pune recently organized a workshop for local school and college teachers, titled ‘Introduction to Machine Learning focussed on application to Biology’.

The author would like to thank Lorena Magana Zertuche for providing the Spanish translation of the paragraph noted above.