CS Bits & Bytes is a bi-weekly newsletter highlighting innovative computer science research. It is our hope that you will use CS Bits & Bytes to engage in the multi-faceted world of computer science to become not just a user, but also a creator of technology. Please visit our website at: www.nsf.gov/cise/csbytes/.

April 29, 2013
Volume 2, Issue 16

Computer Science Enables the Design of Safer Chemicals

Planet Earth

Just because Earth Day has come and gone doesn’t mean we should stop thinking about the environment! This week we examine one of the ways in which computing is helping to make our planet a healthier place.

What makes a chemical harmful to animals or humans? Believe it or not, scientists don’t always know. We only know for sure whether a chemical poses health risks after it has been tested – and unfortunately, these tests are usually done on animals. This testing can be expensive, both in terms of money and in animal life. If only we could predict whether a chemical was likely to be toxic…

Everything around us is made of chemicals, whether man-made or naturally occurring. Some chemicals are a natural part of the environment. Some are good for us. Others are toxic (poisonous) to living things. Scientists have created many useful new "synthetic"(lab-made) chemicals; examples include soaps that help us clean, medicines to fight disease, or materials for building homes.

However, we have not historically known ahead of time whether chemicals are safe or toxic for living creatures and the environment; testing generally must occur after much work and money has gone into making a new chemical. The results of these tests are called "toxic endpoints,"and determine the amount of a chemical that causes adverse effects such as changes in behavior or reproductive capability, or other health effects, including death. Today, scientists are finding ways to avoid both unsafe chemicals and animal testing - with the help of computers!

Chemical Schematic

Schematic illustration of how pre-screening of computed properties can help to estimate the hazards of a given chemical. Image courtesy of ACD/Labs & Sustainability A to Z, LLC.

A variety of software tools exist to help chemists understand fundamental questions of chemistry- how reactions happen and properties of different chemicals. These programs allow chemists to predict and understand various properties of chemicals based on calculations that are far too complicated to perform by hand. This area of study is called computational chemistry.

Researchers are working to inform the design of safer chemicals by comparing computational results to lab tests to look for relationships between computed molecular properties and toxicity. This creates a basis for evaluating newly proposed chemicals before they are even made! Such a predictive approach allows chemists to avoid making molecules with a high probability of toxicity and to focus their efforts on safer chemicals.

One example of this work is that of a multidisciplinary team of scientists. This Designing Safer Chemicals team performed computations for hundreds of chemicals whose effects on algae and other aquatic species had already been determined in a laboratory. They computed the values of a variety of properties for these molecules to see if any of them were correlated to toxicity. They found two! The first is related to how well a chemical dissolves in water, and the second is related to the colors of light that a chemical can absorb. The researchers found that a chemical whose values for these properties fell into a specific range were three to five times less likely to be harmful to the aquatic species. Now, if someone has an idea for a new chemical to make, they can simply compute these properties and get a sense of the toxic endpoint. This allows people who design new chemicals to pursue the ideas that are most likely to be safer for the environment.

Adelina Voutchkova-Kostal and Jakub Kostal

Adelina Voutchkova-Kostal and Jakub Kostal

Who thinks of this stuff? This work is possible because computer scientists, chemists, engineers, and biologists have been working together! Two of the key scientists in this collaboration are Adelina Voutchkova-Kostal and Jakub Kostal. Dr. Voutchkova-Kostal is an organometallic chemist and Professor Kostal is a computational chemist, both at George Washington University. They met while studying in graduate school at Yale University. Now, not only are they scientific collaborators, but they’re married! They live in Washington, DC and enjoy biking, kayaking, hiking, and skiing.

Links:

Read more about Dr. Voutchkova-Kostal’s research on ways to make safer chemicals at: http://departments.columbian.gwu.edu/chemistry/people/111.

Read about computational toxicology research conducted by the U.S. Environmental Protection Agency at: http://www.epa.gov/ncct/.

In honor of Earth Week (last week), did you calculate your impact on the environment?

Try computing your water footprint at http://www.waterfootprint.org/?page=cal/WaterFootprintCalculator, or your carbon footprint at http://www.nature.org/greenliving/carboncalculator/index.htm.

Activity

One of the findings from the Designing Safer Chemicals team is that certain properties of molecules are associated with toxicity. But this does not necessarily mean that these properties cause toxicity. Understanding the difference between association and causation is critical to understanding why variables are related.

Scientists from many fields often examine the relationship between variables. You will have the chance to try this today.

  1. Choose any month from the past year. Based on information you can find on at www.wunderground.com/history/, determine the average temperature outside for each weekday that month. Based on attendance records, determine how many students were present in your class.
  2. Create a table of this data. It might look something like this:
  3. Date Average Temperature Student attendance
         
         
         
  4. Create a graph based on this data.
  5. What do you observe? Are the variables related? If the variables are not related, why do you think this is the case? If the variables are related, would you describe this relationship as an association or a causation? Can you say for sure?

Possible extensions:

For a given month, collect attendance data for the entire school and compare it to the average temperature.