All codes in these files were generated for textmining project by Danielle Lei. Purpose of the code is to generate data visualization for caseOLAP scores of human-heart proteins. Packages and libraries were used include Seaborn, Matplotlib, Numpy, SciKit-learn, and Pandas. Written in Python programming Language.
-
Human heat maps no annot: csv formatted file of caseOLAP scores was input and used. I set a threshold of 0.05 for at least 1 category, so proteins with caseOLAP scores that are greater than 0.05 for at least 1 category was kept. Heat maps were generated by Seaborn package.
-
4_1human heat maps: csv formatted file of caseOLAP scores was input and used. Threshold of 0.1 was set for at least 1 category to ensure better visibility of numerical annotation on the heat map. Heat maps were generated by Seaborn package.
-
4_1human pca: This program was used for PCA analysis on caseOLAP scores. csv formatted file of caseOLAP scores was input and used. PCA analysis was done by SciKit-learn package. Scatter plot and featured-weight bar plot were generated by Matplotlib and numpy libraries.
-
Comparison cross species: This program was used to find proteins which are both present in human and rodent. Data input include caseOLAP csv files (human-heart proteins, rodent-heart proteins, human reference proteome and rodent reference proteome) and corresponding xlsx files downloaded from mapping results on UniRef90. Data processed by Pandas package and numpy libraries. Dataframe of proteins which are present in both human and rodent was generated.
-
bar plot similarity: Data in similarity array were collected from comparison cross species program that was mentioned above. Data in difference array were collected by substracting number of similarity from the total number of clusters which were mapped on Uniref90. Bar plot was generated by Matplotlib library.