A Comparative View Of Applying Linear And Non-Linear Visualisation Approaches To Protein Dataset

Abstract

This novel method enlivened via cartographic maps in the geology area and this technique has been utilized to reintroduce in the visualization space lost data in which the non-linear mapping brings about. The diagnostic measurement of such a bending has been communicated as Magnification Factors, and after that computed then envisioned together as the Cartogram maps We improved interpretability Linear model apply through drtoolbox where cyan circles represent HLAA, red plus sign represents HLA-B and blue square represents HLA-C. Basic purpose behind this study was that previously for large amount of data set, clustering and classifications techniques were used, but through drtoolbox, it is used in MATLAB. The researcher has visualized data for better understanding. This data was aligned in class I HLA-A, HLA-B and HLA-C. Data was available in the form of groups, when it was aligned horizontally then there were 372 rows and 12458 columns. After sorting of data 180 columns remained, Then this data was checked column wise check. The dashes present in the data was replaced by the alphabet displayed at the top of each column. The data coding was done on 12458 rows and data was converted into nominal form. Consensus sequence of data was checked later, the purpose of this sequence is to check the occurrence of each alphabet in a column. The alphabet that was maximum was converted to binary code 1 and remaining were converted to 0. When the data was converted in to binary then models were applied on the data. If the data is in linear form then linear model is better and if the data is in non linear form then non linear model is better, it depends on the results of the data. But in case of this study non linear models showed worst visualization. PCA which is a linear model has showed much better visualization.