Summary

DL has made massive strides in several domains of life sciences and biotechnology, including genomics. CNN architecture is mainly designed for unstructured data. It accepts the input image or a DNA sequence (matrix of size x ) as an input, extracts the features from the image, and does the prediction or classification through a series of hidden layers such as a convolutional layer, a pooling layer, a non-linear fully connected layer, and an output layer. CNNs do not require any separate feature extraction step and automatically derive features from the input data. CNNs have revolutionized the field of genomics because of their incredible accuracy and ability to process unstructured data, which is quite common in genomics.

In this chapter, we have looked at the history of CNNs, what they are, and the different components of CNN architecture. Later in the chapter, we understood how CNNs are being leveraged in genomics for studying complex problems such as gene expression...