Mismatches between order in `variant_consequences` files and the scores we set in our rank models

[fellen31]

E.g. feature_elongation gets score 1, but is the 9th entry in variant_consequences. Further possible inconsistencies: #93 (comment)

[fellen31]

E.g., when using:

info_key = most_severe_consequence

instead of:

csq_key = Consequence
info_key = CSQ

in the rank model, a SV variant in a validation case with feature_elongation&coding_sequence_variant&intron_variant gets 1 point because of the placement of feature_elongation and is not uploaded to Scout, while previously it got 7 because of coding_sequence_variant when not using the most_severe_consequence key, which doesn't care about the placement in the variant consequence list?

[fellen31]

Approximately 1/3 of checked SVs got lower RankScore because of this (some not uploaded to Scout at all), reverting to:

csq_key = Consequence
info_key = CSQ

for LRS for now.