AI Drug Discovery: Unleashing the Potential of Big Data and Machine Learning

First, let’s briefly review some of the basic concepts in the heart of new technologies.

Image credit: gorodenkoff , iStock

Big Data: Volume, Velocity, and Variety in AI Drug Discovery

The term "big data" by itself is more of a marketing nature. It describes an abstract concept of having large volumes of data obtained from various channels in multiple formats, which needs to be arranged in such a way that it can be possible to quickly access, search, update, and analyze it to output useful information.

Today, "big data" is a central strategic concept in most industries, including AI drug discovery, mainly because of the exponential rate of data generation globally — nearly 90% of all data currently available on Earth has been created in the last two years. The computing power required to quickly process huge volumes and varieties of data cannot be achieved via traditional data management architectures using a single server or a server cluster.

Machine Learning: Teaching Computers to Learn in AI Drug Discovery

Machine Learning algorithms are computer programs that teach computers how to adjust themselves so that a human does not need to explicitly describe how to perform the task to be achieved by the computer. The information that a Machine Learning algorithm needs in order to adjust its own program to solve a particular task is a set of known examples.

One of the revolutionary things about machine learning is that it allows computers to learn to perform complex tasks, which are hard or even impossible for humans to describe in "if-then" logic, and instruct. This is called supervised machine learning; it is when the program needs some labeled example data to learn. There are other learning techniques, which do not require a training dataset, for example, learning by “trial and error” — unsupervised machine learning.

Harnessing AI for Drug Discovery: A Synergistic Approach