Once we introduce biopython Structure for representing proteins or amino-acid sequences, we should ensure that this is also a valid input for PSeAAC.

Steps to add this in a programmatic way:

• in PSeAAC, separate transform into a boilerplate transform and an internal _transform which assumes string input
• the boilerplate coerces Structure to str by extracting the amino-acid sequence
• add tests for Structure input

(parked until the pdb / Structure layer is implemented)