Prevent atom alternative locations from ending up in the dataset

DeepRank / DeepRank-Mut

Deep learning framework to predict functional effects of missense variants in human

Apache License 2.0

1 stars 0 forks source link

Prevent atom alternative locations from ending up in the dataset #15

Closed cbaakman closed 3 years ago

cbaakman commented 3 years ago

Some PDB files have alternative locations for certain atoms. These changes make sure that the atom with the highest occupancy is taken, rather than all atoms.

Tests added:

test that of 5MNH chain A residue 153 the atom C is taken, which has the highest occupancy.
test that 5MNH can be preprocessed without error
test that, if a feature does contain NaN values, the entry is not added to the preprocessed data.

cbaakman commented 3 years ago

Thanks Coos, the changes look good. But why is the build failing? Are there other value errors?

I know, that was a surprise to me too. I ran each test individually, but not all at the same time. I'm checking it now..

cbaakman commented 3 years ago

There were indeed still several bugs that caused the tests to fail. The tests do pass now though.

coveralls commented 3 years ago

Pull Request Test Coverage Report for Build 1467968399

65 of 65 (100.0%) changed or added relevant lines in 5 files are covered.
No unchanged relevant lines lost coverage.
Overall coverage increased (+0.7%) to 71.257%

Totals
Change from base Build 1452743238:	0.7%
Covered Lines:	2370
Relevant Lines:	3326

💛 - Coveralls

rgayatri commented 3 years ago

There were indeed still several bugs that caused the tests to fail. The tests do pass now though.

Great! I'll merge this then.