Protein complex identification by graph local clustering and use of Chou´s amphiphilic Pseudo amino-acid features

The intend of this paper is to introduce a new method for protein complex identification. Proteins share an important role by signaling cells. Proteins are three dimensional objects and any types of deformed proteins can cause severe disease, because their function is related to their shape and also the amino-acid sequence they have coded with. Technically speaking similar types of proteins tend to act by forming a group or a complex and if we can find a highly accurate way to identify the complexes we can find a group of proteins that are responsible for something. For this purpose different types of algorithms have proposed but most of them failed to achieve enough precision. Here we will use a new method which uses Chou Pseudo amino-acid as a set of features and we will get a good result even by using machine learning techniques that were supposed as not an effective tool in this field before.