Principal Component Analysis

Principal Component Analysis (PCA) is one of the most famous statistical approach to extract uncorrelated sets of variables that are responsible for the variance in the data.


1. Extract data from the database


Axial ligand

Please select the axial ligands of target hemes. "ALL" includes all kinds of ligand stored in PyDISH other than listed below. The numbers in parentheses represent the number of data of each group stored in PyDISH.




Protein function

Please select the protein function of target hemes. "ALL" includes all kinds of function stored in PyDISH other than listed below. The numbers in parentheses represent the number of data of each group stored in PyDISH.




Threshold of Structural resolution (Å)

Please set the maximum value of structural resolution in Å. More accurate result is expected for lower value, but the amount of data in the extracted dataset becomes smaller.

1.4 8.2

2. [Advanced option] Select atoms used in analysis.

Atoms selected in "Atoms for fitting" are used for the least squares fitting, and atoms selected in "Atoms for PCA" are used in PCA. The atomic labelling in PyDISH (same as in PDB) is shown below.


Atoms for Fitting

At least 3 atoms should be selected.







Atoms for PCA

At least 2 atoms should be selected.









3. Select a target for which the extracted data is compared


Compare for

Graph will be plotted for each group in the target selected here.



Please complete your selection above.