Predicting Suitable Linker for Fusion Protein Using Soft Computing Techniques

Abstract

Protein is a highly complicated substance that occurs naturally, that contains amino-acid residues bounded with peptide bonds. Proteins contain numerous important biological compounds, for example, antibodies, hormones and enzymes. Also, 20 stable amino acids make up a protein. In medicine, proteins are used as antibodies to create vaccines. Initially, proteins are acquired from the cells of plants, animals and microorganisms. There is a remarkable rise in the reproduction of natural proteins through recombinant DNA (rDNA) technology. It further focused on developing “de novo” proteins, which are non-natural and are known as fusion proteins. A fusion protein is a protein formed by combining at least two types of protein domains. Fusion proteins improve bioactivities by the broad range of biotechnological and biopharmaceutical appliances. A victorious building of a recombinant fusion protein needs two indispensable elements: the linkers and the component proteins. The option of the component proteins using a preferred function of a fusion protein manufactured goods is generally relatively uncomplicated. Conversely, Linkers are small peptide sequences. Linkers do not change the operation of individual proteins to which they are attached. Straight fusion of helpful domains with no linker might guide numerous unwanted results, containing misfolding of the fusion proteins, small yield at protein manufacture, or damaged bioactivity. Prediction of a suitable linker for fusion protein is costly because of the prices related to crystallography; electron microscopy also takes more time. In this situation, further soft computing presents numerous feasibilities through creating inexpensive, high-quality results—soft computing methods utilized for fusion protein linker forecast. Here the fuzzy logic is used to forecast the suitable linker for the fusion protein. The experimental results show that the proposed soft computing based linker prediction can predict the suitable linker efficiently.