[ENH] PseAAC feature encoding algorithm

[satvshr]

Since AptaNet (see #13 ) uses PseAAC in its pipeline, a clean implementation of PseAAC is an important pre-requisite for AptaNet.

Todo: add a clean, simplified Python implementation of PseAAC to the repository.
Code: https://github.com/nedaemami/AptaNet/blob/main/feature_extraction.py (part of the official AptaNet implementation)

[fkiraly]

What type is the PseAAC algorithm? Is it pdb -> features?

[satvshr]

Is it pdb -> features?

It is an encoding algorithm that will take the string of the target protein as input and output a feature vector containing its physicochemical properties based on the distance between each amino acid in that protein. Could not answer this question in the meetup, as I had forgotten what pdb was and had to look it up again.

[fkiraly]

ok, so it is pure sequence -> features, right?

Where pure sequence is part of, but not the entire pdb.

[fkiraly]

more software engineering question: I assume you checked that there is no well-maintained package that has this piece? If not, what is the current best implementation(s) in python?

[satvshr]

I assume you checked that there is no well-maintained package that has this piece?

I did! A few packages seemed to implement the algorithm, but I did not look too deep into their source code, as the packages were no longer maintained,

[fkiraly]

can you link them?

[satvshr]

link1
link2
link3

[satvshr]

ok, so it is pure sequence -> features, right?

Where pure sequence is part of, but not the entire pdb.

Did not see this comment, hence the late reply. If by pure sequence you mean a string like "ACDFFKKIIKKLL" then yes.