A Novel Local And Global Similarity Based Feature Extraction Approach For Protein Classification

Each gene acts as a code, or set of instructions, for making a particular protein. They tell the cell what to do, give its characteristics, and determine the way its body works. Each protein has a unique sequence of amino acids. This means that the number and order of amino acids is different for each type of protein. The proteins fold into different shapes. The different shapes and sequences give the proteins different functions, e.g. keratin are a fibrous protein found in hair and nails. If the gene has even the slightest of disorder within its sequence it could lead to an inaccurate order of amino acids

Moreover, the reason

The main focus of this project is reducing the feature extraction time of the system. As a conclusion, it shows that our framework extracts the features from the parse tree very fast. This paper can be further enhanced by using the hybrid classification algorithm to get more accuracy in classification. In this paper, the parse tree is obtained from the PostgreSQL databases and in future, it will get from MySQL databases. To decrease the feature extraction time, fragmented files will be processed in

It is observed from Table 8 and Fig.3 that for each of the five data sets, the highest accuracy is achieved by applying NNGe and Simple Logistics on the feature subset by MLBFSS. The proposed algorithm achieves higher accuracy, lower RMSE, smallest number of selected features compared with other feature selection algorithms. So MLBFSS produces comparatively small number of features with high relevance and is faster as it follows filter method.

A protein has multiple existing structures, these are the primary, secondary, tertiary and quaternary structures which occur progressively. A protein is essentially a sequence of amino acids which are bonded adjacently, and interact with one another in various ways depending on the R group that the amino acid contains. There are 20 different amino acids which are able to be arranged in any given order, thus giving rise to a potential 2.433x1018 (4.s.f) different combinations, and therefore interactions between the various amino acids.

We have used support vector machine (SVM) for classification task. We have used RBF kernel for training the classifier. 10 fold cross-validation is used for determining cost parameter C and best kernel width for RBF kernel function. If we perform classification without any feature selection or feature extraction then the accuracy is 48.99% and 65.82% for AVIRIS and HYDICE image respectively which is very poor and it highly motivates us to apply feature reduction technique. In table II we have shown the classification accuracy for each of the pair of class for PCA, MI and PCA-QMI.

This algorithm was simulated with Matlab. These datasets and the mentioned characteristics are considered and the algorithm of each dataset with different slopes for the activation fumcion of interest were evaluated so that the best slope can be obtained. After running the program for several times and computing the average to obtain the best result, the optimum slope was evaluated for each dataset and the best slopes for Breast Cancer, Diabetes, Bupa, and

The basic building blocks of proteins are amino acids, the biuret reaction tests for protein. A solution of sodium hydroxide is added to a sample then a few drops of copper sulphate solution, if positive – the solution will turn mauve. There are 20 different amino acids and they can be joined in any order. Therefore there can be many different functions. A protein consists of one or more polypeptide chains (a polypeptide chain being multiple amino acids joined together via condensation, producing a peptide bond). Different proteins have different shapes as the shapes are determined by the sequence of amino acids.

In UNIT 7 you learned about the basics structure as well as the basic functions of deoxyribonucleic acid. You learned that deoxyribonucleic acid carries an “instruction manuel” to produce proteins that are responsible for passing traits from your parents to you. As you demonstrated in the previous activity, deoxyribonucleic acid is made of four nucleotides base pairs: adenine, thymine, cytosine, and guanine. They are often abbreviated as A, T, C, and G. The uniqueness of deoxyribonucleic acid is not particularly in just these 4 subunits, but how they are arranged. In this activity you will recreate protein synthesis.

The concept of protein domains and motifs has dominated the first half of this course. Discuss the relevance of protein domains to the following topics:

Amino acids are the smaller building blocks which protein is made of. There are 20 different amino acids. Depending on length, sequence of amino acids and function; each protein has unique different from each other. These amino acids have unique 3D shapes by fold up in complex ways.

For the second part of the experiment, one had to use the knowledge learn from viewing protein molecules in FirstGlance in Jmol to analyze the protein PDB ID: 4EEY. The analysis of this protein was done using the RSCB protein data bank (PDB) at (http://www.rcsb.org/pdb/home/home.do).2

(For example, try entering "myoglobin homo sapiens" to look for the human version of the myoglobin protein sequence.)

In most instances, protein molecules are usually embedded from hundreds to thousands of amino acids. A repertoire of twenty different amino acids, joined in any possible sequence allows the existence of an inconceivably large number of proteins that is infinite in nature.

Different techniques and principles for protein extraction and characterization were demonstrated in this experiment. Various proteins were extracted from different sources: 1.67 g yeast invertase, 1.03 g egg white albumin, and 5.15 g of milk casein. Activity assay for invertase was performed using Benedict’s test and the enzymes inverting action on sucrose was confirmed. Warburg-Christian Method and Bradford Assay were also employed to determine the protein concentration in the albumin and the casein samples. The concentrations for the albumin and casein samples were found to be 0.519 and 0.327 mg/mL, respectively based on Warburg-Christian Assay; and 6.5x10-3¬ and 1.9x10-2 mg/mL