Deal with LINE elements about mpra_selection HOT 3 CLOSED

I think I should filter out the sequences perhaps both before blast and after blast. The ECR11 sequence doesn't give many blast hits outside of primates, so perhaps removing the LINE from that prior to blast will improve the hits in other species.

from mpra_selection.

Okay, I've now (in commit 6418a80) removed repeats from the data pipeline both prior to blasting (I blast with both the masked and unmasked sequence to get better hits) and after. However, I'm going to run into an issue when I get to the final selection calling, since that assumes that the coordinates in the MPRA data match the human sequence, which they no longer necessarily do.

from mpra_selection.

Alright, with the latest bolus of commits, I think I have everything fixed. I ran into an additional issue where different regions were getting masked, so I got around that by propagating the masks after a final alignment.

The coordinate issue I solved by reinserting a homo sapiens sequence with a different name, and having all my scripts know not to mask that.

from mpra_selection.