- database of proteins (should i care about doing collections)
- some portion will have PDB atomic coordinates. but many will exist just as amino acids
- how do i handle states? separate records? when you have variations
- to make life simple just gonna grab the first 1k from FPBase
- viewer of colorpicker
- color picker RGB <> nm wavelength. gives a sorted list by closeness
- picking gives you structure w/ analysis charts
- with trained network, output if there's a fluorescence at certain wavelength/brightness
pipelines (going to use same db for local/prod. don't want to bother w/ multiple envs):
-
FPBase JSON into supabase
-
loop through DB, download PDB files and put into Tigris, then update respective rows with the S3 URL to file
-
agg
-
doi
-
genbank
-
ipg_id
-
name
-
pdb (ids)
-
pdb.0 (ids)
-
seq (amino acids)
-
slug
-
states.0.brightness states.0.em_max states.0.ex_max states.0.ext_coeff states.0.lifetime states.0.maturation states.0.name states.0.pka states.0.qy states.0.slug
-
states.1.brightness states.1.em_max states.1.ex_max states.1.ext_coeff states.1.lifetime states.1.maturation states.1.name states.1.pka states.1.qy states.1.slug
-
states.2.brightness states.2.em_max states.2.ex_max states.2.ext_coeff states.2.lifetime states.2.maturation states.2.name states.2.pka states.2.qy states.2.slug
-
switch_type
-
transitions.0.from_state transitions.0.to_state transitions.0.trans_wave transitions.1.from_state transitions.1.to_state transitions.1.trans_wave
-
uniprot uuid