I have a series of different DNA sequences and structures that I am analyzing in order to determine their efficacy for a certain kind of experiment. Certain regions of the DNA sequences are analyzed based on their percentage of GC content, minimal free energy, etc., and the values of these calculations are assembled into a table. These calculations are carried out using Python.
Now, not only do I want to be able to compare the candidates of DNA sequences, I also want to see which parameter has the most "weight" on the DNA sequences efficacy, so I'm trying to use PCA in order to figure out which parameter is most important for making a DNA sequence the most optimal for my needs.
However, I am not sure how to carry out PCA on the data because the parameters are in two different units– either kcal/mol or in unitless proportions (e.g. the percentage of GC content in a certain region of DNA sequence). Furthermore, the data associated with the kcal/mol have negative value, while proportions are positive values between 0 and 1. Finally, it should be noted that lower values indicate that the candidate has a better chance of being optimal (although with minimal free energy, it is more accurate to say that greater numbers indicate more stability, as in -9 kcal/mol is better than -17 kcal/mol).
Can anybody recommend me any tips for pre-treating this data before carrying out PCA?