Hi everyone, i'm here seeking for opinions on ML algorithms, applied on genetics.

Let's say you have 2 huge datasets:
-one contains a lot of text strings of about 100k characters (only G C A T tho)
-similarly the other contains 10k length strings

some of the strings in a datasets have a relations with some of the other dataset, i do know which ones.

Which ML algorithms would you use to predict relationships given the text strings?

i'm thinking about Autoencoders mainly for dimensionality reduction, RNN or 1D-CNN