Modified Hierarchical Clustering algorithms to Evaluate the Similarities of Growth Factor IR Inhibitors by Using Regression Analysis

In the bioinformatics area it expose an amazing development at the crossroads of biology, medicine, information science, and computer science. The pictures neatly explain that nowadays in this field research is as reproductive in the data mining research. However, maximum bioinformatics research handles with the tasks of identification and classification, tree or network induction from data. Clustering techniques are mostly employed in the sector of information technology, medicine as well as bioinformatics.In this paper, the modified hierarchical clustering algorithms are introduced and applied to orthologous IGF-1R protein sequences and it can produce orthologous clusters of sequences and phylogenetic trees are formed Compared to existing hierarchical algorithms these new algorithms are very efficient, it takes less time to execute and clustering accuracy is also better.Another contribution is acceptable attempt has been made on understanding the role of IGF-1R. The outcome enabled research in extended further to delineate the dependency of Physio-chemical properties, on the activity of inhibitors, and to study the multivariate regression analysis on a set of 87 IGF-1R inhibitors are dependent variables and some of independent variables resulted in F-test: 8.812, r value: 0.794 and r2 value of 0.631, respectively. The data set was introduced for the presence of outliers by calculating the leverages and standard residuals and finally 6 compounds were eliminated. A new regression model was attempted 76 compounds training set and 5 compound validation set. A Regression plot is obtained and justifies the predictive ability of the regression model. Finally, the designing or screening compounds libraries for new analogues should enhance the inhibitory activity against IGF-1R.