Category Archives: Uncategorized

Track Your Stress with Sweat

Do you hate having your blood drawn for your lab tests? Could there be a non-invasive way to obtain your lab results?  Perhaps, sweat samples could be used to measure our health status instead.

Sweating is a naturally occurring process, whether it is from exercising or getting nervous on a test. Although sweat can be perceived as wet and smelly, sweat contains various types of biomarkers, such as the stress hormone, cortisol. Since excessive stress can contribute to various health problems, such as high blood pressure, could we use cortisol in our sweats to monitor our stress levels in real-time?

In a recent study, Sekar et al. has developed a wearable electrochemical sensor that can measure cortisol in sweat. The researchers has integrated iron (III) oxide (Fe2O3) in conductive carbon yarn (CCY) to make a semiconductive platform. After that, the platform is coated with antibodies (anti-Cmab) in an electrochemical apparatus, which would make the sensor specific to cortisol. The final product would then become a Fe2O3/CCY immunosensor. The purpose of the study is to see if they can use CCY as a suitable platform for biosensors when monitoring sweats.

Adapted from Figure 1b in the Sekar et al (2019) paper. The black rectangle is the CCY with iron (III) oxide (orange spots). The green cylinder is the electrochemical apparatus. Licensed under a Creative Commons Attribution 4.0 International License

The researchers were able to test the sensor’s detection ability with different concentrations of cortisol. According to Figure 8b below, the line graph shows a linear relationship between the electrical current response from the Fe2O3/CCY immunosensor and the logarithm of cortisol concentration.

Adapted from Figure 8b in the Sekar et al (2019) paper. Each data point with error bar is the result from three successive experiments. Licensed under a Creative Commons Attribution 4.0 International License

The researchers also tested the sensor with real sweat samples from participants after performing cardio exercise. In the bar graph below, the error bars in the pink bar gives the RSD or relative standard deviation of 3.403%, 3.874%, and 4.064% from sweat sample 1, 2, and 3 respectively. These RSDs show small variations in averaged results from three successive experiments when testing with the Fe2O3/CCY immunosensor. According to the paper, the bar graph below shows a correlation between the two methods: the CLIA (chemiluminescence immunoassay) and their Fe2O3/CCY immunosensor. As a results, using CCY may be a possible choice for designing a biosensor that monitors cortisol in sweats.

Adapted from Figure 11 in the Sekar et al (2019) paper. Each pink bar with error bar is the result from three successive experiments. Licensed under a Creative Commons Attribution 4.0 International License

In addition, there are other similar studies that focus on wearable sweat sensors, such that they can transmit data to your phone, and diagnose cystic fibrosis. Therefore, sweat sensors are potential non-invasive diagnostic tools, which may lessen the burden on more invasive blood samples to measure our health status.

References

Stress and Heart Health. https://www.heart.org/en/healthy-living/healthy-lifestyle/stress-management/stress-and-heart-health (accessed Mar 26, 2019).

Sekar, M.; Pandiaraj, M.; Bhansali, S.; Ponpandian, N.; Viswanathan, C. Carbon fiber based electrochemical sensor for sweat cortisol measurement. Scientific Reports 2019, 9, 1-14. https://doi.org/10.1038/s41598-018-37243-w.

Stephanie, Relative Standard Deviation: Definition & Formula. https://www.statisticshowto.datasciencecentral.com/relative-standard-deviation/ (accessed Mar 26, 2019).

Geddes, L. Wearable sweat sensor paves way for real-time analysis of body chemistry. http://www.nature.com/news/wearable-sweat-sensor-paves-way-for-real-time-analysis-of-body-chemistry-1.19254 (accessed Mar 26, 2019).

Dusheck, J. Wearable sweat sensor can diagnose cystic fibrosis, study finds. http://med.stanford.edu/news/all-news/2017/04/wearable-sweat-sensor-can-diagnose-cystic-fibrosis.html (accessed Mar 26, 2019).

 

Image

Does climate change result in an increase of wildfires?

As the temperatures get warmer and the days start getting longer, you may be excited for the summer season to start. However, the arrival of summer also means dryer weather, water shortages, and wildfires.

Poor air quality in Vancouver due to the raging wildfires.         Photo Source: flickr

Wildfires can occur anywhere, but are most prevalent in the forests of Canada, the United States, Europe, and throughout the vegetated areas of Australia and South Africa. These wildfires are capable of destroying entire areas of up to 2.5 million hectare per year in Canada and can travel at speeds of 23 kilometers per hour! The main causes of wildfires are lightning and humans (surprise!), with lightning caused fires making up 45% of all fires and 55% of all fires being connected to human activities.

Historical wildfires across Canada. Photo Source: Natural Resources Canada

Although lightning caused wildfires occur naturally every year and are essential for the environment, wildfires should still be a concern for all of us. The smog that visits Vancouver every summer seems more severe year after year, could climate change be increasing the occurrence of wildfires?

According to science, climate change does result in an increase of wildfires. Climate change is the result of human activities, such as, fossil fuel burning, which produces large quantities of carbon dioxide. Just like methane and ozone, carbon dioxide is also a heat-trapping molecule, which you might know as a greenhouse gas. With greenhouse gas concentrations increasing, the heat radiated from earth cannot leave earth’s atmosphere. Over 90 percent of this trapped heat accumulates in the ocean and as a consequence, ocean heat contents rise and cause increases in global ocean temperatures. The increase in global heat content also causes ice to melt and sea levels to rise.

Global ocean heat content (OHC) for the top 2000 m of the ocean. Uncertainty estimates are shown in pink. Source: Science Advances

Aside from an increase in global ocean temperatures and higher sea levels, temperatures on land are also raised due to the trapping of heat by greenhouse gases. This results in more droughts and dryer weather, which are perfect conditions for wildfires to pick up.

So, other than knowing the reason for an increase in wildfires, how can we do our best to prevent wildfires from occurring?

First of all, always obey fire bans and signs that indicate the wildfire danger ratings. Next, never leave a campfire unattended & completely extinguish the campfire when leaving the area. Furthermore, any smoking material should be properly disposed of and most importantly, be sure to report any signs of wildfires.

Can Weather Affect Our Moods?

You wake up in the morning and get dressed for the day. But the sky’s overcast and you can’t even get a glimpse of the sun. It’s already pouring outside and you utter to yourself, “This day’s off to a great start.”

Picture of a Rainy Day; Source: Max Pixel

Living in a city like Vancouver, where it rains for nearly half the year, you may start to wonder whether the weather actually has an effect on your mood. Is this merely a myth? Or does the idea have some merit to it?

From this paper published in 2011, the researchers came to the conclusion that it depends. Their results show that some people’s state of mind can be influenced by the weather. While for others, there was little to no effect.

They quantified this influence by focusing on three distinct mood indicators: happiness, anxiety, and anger. To measure this, the Internet version of the Electronic Mood Device, the Daily Mood Scale was used. Klimstra, et al. defined four groups of people – each affected differently by weather. They labelled these groups as: Summer Lovers (SL), Unaffected (UA), Summer Haters (SH), and Rain Haters (RH).

Graph Showing percentage of each type of individual; adapted from Klimstra, et al.

From their results, they found that nearly half of the individuals tested were in the UA group – meaning that the weather’s influence was minimal. As suggested by the writers of the paper, this could explain why previous studies concluded that the weather had an insignificant effect. However, the rest of the participants displayed changes in happiness, anxiety, and anger in the weather conditions that were tested – weather did have an effect on them.

Correlation results from their study. Source: Klimstra, et al.

The table above shows the correlation between sunshine, temperature, and precipitation and their effects on the mood indicators. Positive numbers indicate a positive correlation, whereas negative numbers indicate the opposite. The correlation ranges from -1 to +1; a number closer to the extremities (±1) indicates a stronger link. The number of people (of each group) is indicated by n. The probability value (p) represents how statistically significant the results are – the smaller the p-value, the higher the level of significance.

Summer day in Biei, Hokkaido, Japan. Source: Reginald Pentinio

You can see why they decided on these labels for each group. For example, individuals of the Summer Lover group were characterized as being happier, being less anxious, and being less angry on higher temperature and sunny days. However, precipitation caused reduced levels of happiness and increased levels of anxiety and anger.

It’s safe to say that the weather does have an effect on at least half of us.

This blog post’s focus was mainly on the study by Dr. Klimstra. The following video mentions some other ways the weather can affect us.

Ketamine as an antidepressant. Is that oK?

Since its development in 1962, ketamine has been used primarily as an anesthetic for veterinary procedures. In recent years, however, research investigating its use has extended to psychiatry, with evidence supporting ketamine as a viable treatment for Major Depressive Disorder, colloquially known as depression.

https://www.researchgate.net/figure/Chemical-Structure-of-Ketamine-5_fig2_320345763

Figure 1. Chemical structure of Ketamine

The most commonly used antidepressant drugs are tricyclic antidepressants and selective serotonin reuptake inhibitors (SSRI). Upon daily administration, these drugs relieve depression, but only after approximately 3- 6 weeks. Moreover, for those with treatment-resistant depression (TRD), SSRIs prove to be of little benefit. Remarkably, current studies suggest that ketamine improves symptoms within 30 minutes, with therapeutic effects for even TRD patients.

An article published in JAMA elucidates the potential of N-methyl-D-aspartate (NMDA) ketamine for the treatment of TRD: in a preliminary study involving eight subjects with depression, Zarate et al determined that a single dose of NMDA ketamine resulted in a rapid but short-lived antidepressant effect. In a subsequent double-blind randomized clinical trial, subjects received intravenous infusions of ketamine hydrochloride or midazolam as placebo. The participants used the Hamilton Depression Rating Scale (HDRS) to measure the changes in drug efficiency. As seen in the linear mixed model, in Figure 2, the difference between ketamine and placebo treatment over 9 points from baseline to 7 days were examined with standard error. Within 110 minutes after the injections, participants receiving ketamine showed significant improvement in depression compared to subjects receiving placebo (with P<0.05).

Figure 2. Changes in the 21-item Hamilton Depression Rating Scale (HDRS)

The implications from this study and the many other breakthrough studies have not gone unnoticed. The development of chemical variations of ketamine has shown that the drug is a powerful tool that can allow people to live life to the fullest potential. In fact, on March 5th, 2019, the Food and Drug Administration (FDA) approved Esketamine, a nasal spray formulation derived from ketamine, for TRD. Targeting the brain’s glutamate pathway, Esketamine is the first drug in thirty years to be approved with a new mechanism of action for treating depression. Of course, the FDA approval of Esketamine does not negate the lingering caveats and concerns relating to its abuse. Nonetheless, even as studies continue to investigate Esketamine’s adverse effects, many physicians remain optimistic that it may become “the biggest breakthrough in depression treatment since Prozac.”

Figure 3. Esketamine, which will be marketed under the trade name, Sparavato

In the following video, Dr. J. John Mann at Columbia University highlights the importance of the approval of Esketamine and the potential risks involved.

-Brina

Is it actually 100% oregano?

Have you ever wondered what is in the food you eat? This pizza may contain additional ingredients that you may not be aware of.

According to Canadian Food Inspection Agency (CFIA), food fraud is an emerging global issue. In fact, food fraud “may cost the global food industry $10 to $15 billion per year”. Examples of food fraud may include substitution/addition of ingredients or tampering/mislabeling of food packages, and selling these inferior products at a higher price for profit. Food fraud is problematic; therefore, it is crucial that CFIA and the food industry combat food fraud to protect consumer safety.

However, in 2016, there has been a report of adulterated dried oregano in Australia. Some brands that declare “100% oregano” only have 33% – 50% of actual oregano. The remaining percentage could contain additional olive and myrtle leaves as fillers. The presence of olive and myrtle leaves can pose a health risk, because it can carry a higher amount of pesticides, which can contaminate the dried oregano. Therefore, it is important to find a way to detect these fillers, so that they can be eliminated from the market.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Recently, a paper from the journal of Food Chemistry published in 2019, suggests that GC-MS (a common instrument in a Chemistry lab) can be used to detect and measure the amount of pesticides in adulterated oregano samples. By identifying the most predominant pesticides in adulterated oregano, the pesticides can be used as potential markers for identifying adulterated oregano.

But how does GC-MS work? In the “GC” part of the instrument, the pesticides travel through the column, in different speeds, based on its unique chemical properties. Once all of the pesticides are separated, they go through the “MS” part of the instrument, where they get fragmented by a beam of electrons before it travels through the mass analyzer and reaches to the detector for data collecting (see image below).

A schematic of the GS-MS instrument. Detector is attached to the right side of the mass analyzer (not shown). Cwszot, KkmurrayCreative Commons  Attribution 2.5 Generic (CC BY 2.5), Electron ionization GC-MS.png

As a result, pesticides (cyfluthrin (sum), cyhalothrin lambda, and pyriproxfen) are present in higher quantity in the 34 adulterated oregano samples than in the 42 genuine samples. Therefore, cyfluthrin, cyhalothrin lambda, and pyriproxfen could be used as potential markers for detecting adulterated oregano.

Graph from the research paper. Click on the image for high-definition. Drabova et al., Creative Commons Attribution 4.0 International (CC BY 4.0),  Adapted from Figure 5 in Food fraud in oregano: Pesticide residues as adulteration markers

In conclusion, it is possible to identify the adulterated samples by using a chemical technique to stop food fraud. Although CFIA and food industries work to protect consumers from food fraud, CFIA suggests a few ways for consumers to identify food fraud.

But as for me, I will stick to growing my own oregano in my backyard.

Updated: March 28, 2019 

Reference:

Canadian Food Inspection Agency. The CFIA Chronicle. http://www.inspection.gc.ca/about-the-cfia/the-cfia-chronicle-fall-2017/food-fraud/eng/1508953954414/1508953954796 (accessed Mar 08, 2019).

Canadian Food Inspection Agency. Food fraud. http://www.inspection.gc.ca/food/information-for-consumers/food-safety-system/food-fraud/eng/1548444446366/1548444516192 (accessed Mar 08, 2019).

Canadian Food Inspection Agency. Types of food fraud. http://www.inspection.gc.ca/food/information-for-consumers/food-safety-system/food-fraud/types-of-food-fraud/eng/1548444652094/1548444676109 (accessed Mar 08, 2019).

The Sydney Morning Herald. Food Fraud: Popular oregano brands selling adulterated products. https://www.smh.com.au/business/consumer-affairs/food-fraud-popular-oregano-brands-selling-adulterated-products-20160405-gnygjo.html (accessed Mar 08, 2019).

Drabova, L., Alvarez-Rivera, G., Suchanova, M., Schusterova, D., Pulkrabova, J., Tomaniova, M., . . . Hajslova, J. Food fraud in oregano: Pesticide residues as adulteration markers. Food Chemistry. [Online] 2019, 276, 726-734. doi:10.1016/j.foodchem.2018.09.143 (accessed Mar 08, 2019).

Canadian Food Inspection Agency. How food fraud impacts consumers. http://www.inspection.gc.ca/food/information-for-consumers/food-safety-system/food-fraud/how-food-fraud-impacts-consumers/eng/1548444986322/1548445033398  

The Link Between Stress and Technology

Whether it is stress over a failed exam, or feeling stress when put in a new environment, we have all experienced stress before and know that being stressed is not a good feeling.

Figure 1. Participants rate their stress level on a scale of 1-10, where 1 equals “little or no stress” and 10 means you have “a great deal of stress”. Photo source: American Psychological Association

Although stress usually has a bad connotation associated with it, there is also positive stress. For example, positive stress can motivate you and help you complete tasks more efficiently. However, high levels of stress can lead to anxiety, depression, high blood pressure, and other chronic illnesses. Stress levels found in humans have increased drastically over the years and may be a concern to a high percentage of the human population. There are many factors that can explain the increase in stress over the generations, but let’s focus on one that we are all familiar with: technology.

Could the increase in stress over the years be linked to technology use?

Let’s first look at smartphone ownership over the recent years.  In 1992, the first smartphone, the Simon Personal Communicator by IBM, was introduced and 15 years later, the first iphone was produced. The popularity did not start immediately and only began increasing around 2011, where approximately 35% of U.S. adults owned a smartphone. The percentage quickly ascended over the next 6 years to 77%. With more advances in technology, people are now overly dependent on their smartphones. But, who can blame us for being overly dependent on our smartphones? Not only can it take high quality videos/photos, but you can even pay with smartphones nowadays – it’s all just one tap away.

                             Figure 1. Smartphone Ownership                                Adapted from: Pew Research Center – Internet & Technology

So, what evidence is there that proves that technology adds stress to our lives? Well, sleeping patterns can be easily affected by technology, one second you’re getting ready for bed, and the next second you are asking your friends whether the dress is black and blue, or white and gold and then you realize it’s way past your bedtime. The constant distractions our smartphones present make us less efficient when completing tasks as simple as going to bed at an appropriate time.

Image result for blue light screens

Photo source: Lifewire

Furthermore, the blue light emitted from our phone screens can reduce melatonin production, which reduces your sleep quality. When sleep quality is reduced, one then becomes less resilient and stress levels and anxiety increase.

In conclusion, it is evident that the accessibility and convenience of technology can negatively affect our living qualities by causing stress levels to rise. As stated above, sleep is easily affected by technology use and sleep is crucial for out bodies to function properly. Therefore, although technology can be very handy, one should be aware of the effects of technology on your stress levels.

Watch the following video for more information on the effects of technology:

The Age of Misinformation

Misinformation is false information that is spread, regardless of whether there is intent to mislead or not.

With the internet quickly gaining popularity near the turn of the century, information sharing had become easier than ever before. News traveled quickly as news-agencies could publish online and social media could spread it. Anybody could share their thoughts on websites. The Information Age was here.

Social Media Platforms that Americans use for news; adapted from American Press Institute.

Compared to the past, the internet offered a much faster medium for information. This was beneficial because plenty of info became easily accessible to the public. However, both experts and people who claimed to be the former could share their knowledge.

With the speeds at which information could now move, misinformation could also spread quickly. Take the Boston Marathon bombing incident as an example. On social media websites (like 4Chan, Reddit, and Twitter) there were internet sleuths hard at work identifying the Boston Bombers. Even credible news-agencies were racing to report the information without proper verification. This caused more harm than good; false information was given credibility. One family had to remove a Facebook page that they had put up to find their missing son.

Social Media Platforms. Source: kisspng

 

On the internet, everyone has a voice. It is simple for anybody to find a community that shares a point of view that they agree with. Sometimes, this idea may not be correct. One only needs to look at vaccines to encounter this scenario. In the past few decades, some people have begun to reject vaccinations and declare vaccines to be the cause of autism spectrum disorder. While it is true that the rate of diagnosis has increased in this time period, this is likely due to advancements in autism research.

Autism prevalency overtime (diagnosis was 8 years after year of birth); adapted from CDC.

Unfortunately, the idea of this link quickly gained traction after Andrew Wakefield published his research about the relationship between the usage of the measles, mumps, rubella (MMR) vaccine and autism. Eventually, he had his paper retracted and he was barred from practicing medicine in his country.

Vaccine Clipart. Source: Clipart Panda.

Even after many studies and investigations found no link between autism and the MMR vaccine, there were still people who believed in vaccines being the cause. Even after proof of his conflict of interest- his patent, there were still people who believed him. And at present, people who oppose vaccines can find like-minded individuals to share and cement their beliefs. Once an idea has been accepted by an individual, it becomes highly resistant to correction.

Example of misinformation and the inability to correct; Source: Chris Meserole (Law Fare Blog).

Further complicating the situation, everybody also tends to prefer to read content that relates to their narrative while ignoring other ones. Whether it be a news article, a comment, or even a blog post like this one, one must remain diligent about the content they are reading.

Fast, Faster, Fastest

Fast, Faster, Fastest

List of the world’s fastest 100 meter sprinters. Shorter time is better. Data retrieved from rankings.com

From horses to cars, trains to spaceships, many aspects of the world are happening in smaller time frames. For example, from Carl Lewis’ 1984 Olympic performance to Usain Bolt’s 9.58 s record at the Beijing Olympics. However, while sprinting has been slowly reaching its plateau, computational power has been exponentially growing.

Exponential growth of computational power over time

Modern day computers run at incredible speeds thanks to the even faster networks that have been developed today. From vacuum tube transistor to the 10 nm technology rolled out from Intel. Moore’s law is based on an observation by Gordon Moore, which found that the number of transistors that can fit inside an integrated circuit doubles every two years. 

Since 2017, Moore’s law has shown signs of stagnancy. It should have been obvious that at some point, this prediction would no longer hold up simply due to the physical limitations of shrinking the transistor size. With contemporary technology, the resources of constructing a CPU (Central Processing Unit) with 14 nm transistors, which is equivalent to 70 Silicon atoms, is entirely feasible. When transistors begin to approach the size of electrons, not only is it harder to trap electrons which are what makeup electrical currents inside the computers and dictate the 0’s and 1’s bit of information stored on the computer, manufacturing these smaller transistors also poses a problem as well. CPUs which essentially the brain of the computers, are made up of dies cut from large silicon wafers which need to made from pure silicon (99.9999%). The problem is to make the patterns on the die, one needs to have a laser source that is already much smaller than the wavelength of purple light: 400 nm. Lasers with shorter and shorter wavelengths are being used and eventually similar to microscopes the wavelengths of light are not longer able to be manipulated smaller and other sources will have to be used.

While the future seems uncertain on the fate of future computers, it is also amazing to think that computers that were used to launch mankind onto the moon are now dwarfed by the power of electronic devices that fits into our pockets. For the vast majority of consumers, it is not how fast the computer runs that dictate our productivity but how we decide to utilise it.

References:

(1)
(2)
From Sand to Hand: How a CPU Is Made. Geek.com, 2009.
(3)
Jurvetson, S. Moore’s Law over 120 Years; 2016.
(4)
Top Ten Fastest 100 Meter Sprinters in History https://www.rankings.com/sports-track-sprinters/ (accessed Mar 2, 2019).

Battery for future electric cars

In recent years, the car industry got a huge revolution for the blossoming of electric cars. More and more people put attention on electric cars for its quietness while running and amazing acceleration ability. Also, since electric cars are driven by electric motor, means no refuel on the gas station thus no direct CO2 emission, this environment-friendly property has been greatly spread by major media.

Source: C.-X. Zu & H. Li Energy Environ. Sci. 4, 2614–2624 (2011)/Avicenne

 

However, there are two major problems dragging the wide application of electric cars back. First, the cost of making the batteries for electric cars are extremely high. Winfried Wilcke, heads of IBM’s nanoscience and technology division got interviewed and said: “battery packs for electric cars cost more than $500 kWh−1“. While, in comparison, it is reported by The Union of Concerned Scientists, the cost of making electric cars are comparable to gasoline-powered cars if the cost for battery packs is between 125 and 150 kWh−1. Second, the range of electric cars from one charge is significantly smaller than gasoline cars. The range of electric cars is averaged at around 150mi while gasoline cars are around 400mi.

An electric car made by Tesla Photo by Jp Valery on Unsplash

 

For solving those problems mentioned above, research groups around the world are trying different solutions and developing new batteries. Magnesium-ion battery and  Sodium-oxygen battery are two of the approaches scientists working on. The Magnesium-ion battery uses magnesium ion that can carry the double amount of electrons of lithium carries and it migrate in pairs. In total, magnesium ions can carry four times more charges that currently used lithium ions.

For the sodium-oxygen battery, it can only provide half of the range supported by lithium-oxygen battery but 5 times more than the lithium-ion battery. Also, sodium is cheaper than lithium, therefore the cost of making battery pack to provide the same amount of range can be cut down to as low as 1/5 as the lithium-ion battery.

Adapted from Tao Liu, Gabriella Bocchetti and Clare P. Grey

As mentioned above, the lithium-oxygen battery has an amazing energy density to make the long-range electric car become reality, but as reported by a group in the University of Cambridge, the impurities in the air can clog the electrode of the lithium-oxygen battery and this damage the battery after few dozen charges. But researches are investigating the reasons behind it and finding solutions to solve it.

Is Machine Learning the Future of Technology Development and Chemistry Research?

The ability for scientist to develop new drugs for everything from rare diseases to headaches is often reliant on precedent and systematic investigations. These methods are often costly and time consuming. Similar problems arise in development of new materials that may enhance our energy production. Our limited ability to rationally design materials  hampers their development. This leads to reliance on our ability to recognize the trends and behavior of already existing materials. However, what if we could amplify the ability to recognize patterns beyond human limits? Machine learning answers this problem.

A graph depicting the general algorithm machine learning follows. Source: Wikimedia Commons

While machine learning is a form artificial intelligence, our jobs are safe. Machine learning is the use of statistics and the power of computers to predict results or identify trends in data. The general method relies on the input of “training” data which is analyzed using statistics. After developing a model, information may be inferred from new data the computer encounters.

-Video Source: Google Cloud Platform educational AI Adventures Series on YouTube by Yufeng Guo in 2017.

Large technology companies have recognized the advantage of integrating machine learning into technology development. Google is one example that has successfully introduced it. Gmail uses machine learning to service 1.5 billion active accounts. They claim to detect 99.9% of phishing and spam mail from entering the user’s inbox. However, machine learning is not limited to technology companies. Chemistry researchers have quickly adopted it.

Total Number of Chemistry Publications with “Machine Learning” in Title

Starting in 1969, the first chemistry journal article with “machine learning” in the title was published. By combining machine learning with a common technique called mass spectrometry, Peter Jurs at the University of Washington was able to determine chemical composition of “unknown” chemicals using the input of 348 unique patterns as “training” data.

More recently there has been an almost exponential increase in the number of chemistry publications applying machine learning. In the last two years approximately 6 times as many publications were made than in the past 48 years. Tommi Jaakkola, a Professor of Electrical Engineering and Computer Science at MIT said at a consortium about implementing machine learning in the pharmaceutical industry: “by marrying chemical insights with modern machine learning concepts and methods, we are opening new avenues for designing, understanding, optimizing, and synthesizing drugs.” The materials science community has also seen integration with the development of novel long chained molecules called polymers for photovoltaics by scientist at Osaka University. Shinji Nagasawa, the lead author explained the importance: “there’s no easy way to design polymers with improved properties. Traditional chemical knowledge isn’t enough. Instead, we used artificial intelligence to guide the design process.”

Solar cell efficiency over years showing a substantial increase. Source: Wikimedia Commons

While machine learning is not the solution to all chemical problems or spam mail, it is being widely accepted by the scientific community and technology industry for good reasons. Even with limitations, it’s effectiveness across a wide array of industry and research emphasizes the role it may play in the future of research and development.

—Jonah

References

  1. Graph-powerd Machine Learning at Google. Google AI Blog. https://ai.googleblog.com/2016/10/graph-powered-machine-learning-at-google.html (Accessed Feb 28, 2019).
  2. Jurs, P.C.; Kowalski, B.R.; Isenhour, T.L. Computerized Learning Machines Applied to Chemical Problems: Molecular Formula Determination From Low Resolution Mass Spectrometry. Chem. 1967, 41, 21-27.
  3. Machine Learning, Materials Science and the New Imperial MOOC. Imperial College London. https://www.imperial.ac.uk/news/187054/machine-learning-materials-science-imperial-mooc/ (Accessed Feb 28, 2019).
  4. UBC Summons. University of British Columbia.