While there is immense progress done in terms of scientific research, some aspects of biological research are very hard and time-consuming. Imagine you are finding out a suitable drug for a disease. To begin with, you must know what 
causes the disease, not the triggers but the proteins within the body which stop functioning properly thereby causing the disease. Now, this is called a target for developing drugs to combat that disease. The next step would be to find out drugs that could act on that protein in a specific manner. For example, let us consider the protein to be a lock, the drug should be like a key that can fit into it correctly and help open or close the lock.
Nature, as well as chemistry, has an abundance of molecules that could be used as drugs. You would be surprised if I told you that most of the medicines, we consume, are either molecules found in nature or inspired by them. Given this scenario, in a lab, it would be next to impossible to conduct experiments of such
kind on a large scale. It would take years or probably decades to complete these experiments on just one target, much like Harry Potter trying to find the one right key among the many winged ones flying around, to open the door to the dungeon.
With this major obstacle, some smart scientists decided to explore using computers to help overcome this situation. They trained computers to predict the structures of proteins by teaching them basic principles and rules governing protein structures. They taught the machines to decode what could be a potential ‘lock’ site to design the keys accordingly. With this done, they trained the computers to investigate the possible structures of the small molecules based on data obtained experimentally and then tried to see if the ‘keys’ thus made would fit into the protein ‘locks’. This was something 
revolutionary in biology and made work a lot faster and simpler for the scientists involved in drug discovery. This was the simple beginning of computational biology, a branch of science that integrates two very important aspects of science, and together they formed an interdisciplinary branch of science that has made discoveries much easier.
Computational biology has also helped in analyzing sequences of DNA, RNA
and in the prediction of structures of proteins. These tools are also used to generate huge databases or libraries of biological molecules or compounds from which one could with ease find out something they are looking for. It could also help to find out closely matching relatives of a newly discovered protein which can help in identifying its function. Designing experiments, analyzing the data obtained, shortlisting the most suitable molecules (the top 100 or so) to save time as well as costs in a project, are just some of the advantages of this branch of biology.
What I have listed out is just the tip of the iceberg and with advances in Machine Learning, Artificial intelligence, and Data Science, the potential is much much more. Recently using AI, structures of almost 98% of the human proteins have been predicted, a task which would have taken centuries of hard-core experimentation, done in a matter of a few years.
To read more on this branch of biology visit:
http://cbd.cmu.edu/about-us/what-is-computational-biology.html
