What is computational biology?
Computational biology is recreation of life in computer, using mathematical modeling, simulations and data analysis. Computational biology, an interdisciplinary science, has quite a wide spectrum and highly involves biology, organic chemistry and computer science.
How did we come up with computational biology?
Computational biology was born as science field in 1990’s with the developing computer technology. Why did someone want to study organisms in computer instead of living environment? Studying the flora, fauna, internal organs and life cycle of an insect, metabolism and decision mechanism of a single cell, or even the organic molecules in their own niche provides a lot data. However, we can do similar observations or tests in silico as well. Understanding biological systems requires computing as much as observation. We make use of computational biology for all systematic approaches. For instance by modeling the life cycle of a bird species, which is about to get extinct, you can calculate how many birds you need to release to save the species. Or, if you model the underwater fauna of a lagoon, you can predict the effect of an invading algae species and test the possible precautions in silico rapidly. There is not big difference between modeling methods in computational biology. A model similar to the one used for invasive algae can be applied for invasive tumor cell (metastasizing malign tumor cells) as well. These kinds of models are very important for developing cancer therapies. If we deal with behavior of the organic molecules instead of organisms, we can model their movements and reaction in computer.
Computational biology, born as simple calculations 30 years ago, has turned into a frequently used and popular field. 2013 Chemistry Laureates Martin Karplus, Michael Levitt, and Arieh Warshel won the Nobel Prize for their work in computational biology, titled “Development of Multiscale Models for Complex Chemical Systems“.
What can we do with the computational biology?
We can model many elements of life in computer. Having these models will allow us predict the behavior of the systems. For example, several fields of science study human brain from different perspectives (imaging, physiology, cell biology, neurochemistry, psychiatry etc.). There is an effort to model the human brain using computational biology, which brings data from different fields together. Such a model would be extremely helpful for neurobiological studies.
The branch of computational biology applied in our department is molecular modeling. Organic molecules (lipids, DNA, protein etc.) can get billions of different shapes in 3D space. But in physiological conditions they have the most stable few shapes. Knowing the shape (3D structure) has a vital importance since it specifies the function too. Starting from the genomic sequence, we can predict the primary, secondary, tertiary and quaternary structure of proteins (figure below). We can compute the 3D shape and dynamics (movements) of the molecule using the chemical properties, interaction with water, X-ray information etc. Understanding molecular movements and interactions is one of the major steps humanity has to achieve to understand life.
For example this video, prepared with the help of biochemistry, structural biology and computational biology tells a lot about how the protein are packaged in neurons.
Protein simulations give clues about many disease mechanisms, which could be target for treatment. As an example, you can check the video, which describes the fusion of Dengue virus.
Computational biology in business
The application of assigning 3D structures of enzymes, which are naturally proteins, is the drug research. With the aid of quantum chemistry, we can compute how and how strong would the drugs interact with the target enzyme, and if the drug will pass through cell membrane. We can do this for novel drug candidates, as well as well known drugs. The drug candidate molecules can be screened in silico within few days, which would take years to synthesize and test in wet lab. We can also design a totally new drug, against a novel target enzyme, specifying the target region, specificity, binding efficiency etc.
Drug design is a long and laborious work. Over 10 billion drug candidate molecules must be screened. Only few of them pass the chemical tests, animal tests and clinical tests to be presented to FDA. For FDA to accept a molecule as drug, it should have high activity, minimum side effect and should be able to reach the target through membrane. There are so few novel molecules to fill the requirements. FDA accepted only 18 novel molecules as drug in 2013. Synthesizing 10 billion candidate molecules chemically and testing on animals is practically impossible due to high cost and time. With the latest technology, computational biology greatly reduce the cost and speed up the process, by making a pre-selection to the candidates. Not only this, but we can also calculate the interaction of the drug molecule with the other molecules, and possible side effects with computational biology. This prevents many improper candidates to be tested on animals.
The side effects of the drugs should be screened perfectly. Even tiny chemical interactions that miss from attention might lead to terrible consequences. In 1960’s, Thalidomide was commonly used against the morning sickness in pregnant women. Its side effect on fetus extremity development was noticed only after many handicapped babies were born (photo down). These days, such side effects are eliminated at the computational level.
Similarly, cancer drugs have been greatly improved for the 40 years. Today’s chemotherapy agents are incomparably more effective with much less side effects. Computational biologists are highly seeked scientists by the pharmaceutical companies and institutes of basic science, due to their role in drug design.