A static class that provides optimization hints for memory or performance handling of sequence data.
If you are working with genomic sequence data and the use case is sub sequence then the sequence proxy loader
implementation may do file reads for each sub-sequence request instead of loading everything into memory.
If a large collection of protein sequences is being loaded from a fasta file but only a few sequences will be selected
then the file loader could create a sequence proxy loader that retains the offset in the file for each sequence
and it is loaded when the sequence data is requested. This way you could load a pointer to all sequences but delay
loading until the user actually needs the data. The sequence loader could also have an internal sequence management
algorithm that goes through and returns sequence data freeing up memory where a future request for sequence data will
then be reloaded.