Science Score: 54.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
✓CITATION.cff file
Found CITATION.cff file -
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
○DOI references
-
✓Academic publication links
Links to: zenodo.org -
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (8.2%) to scientific vocabulary
Repository
Deep learning for estimation of kcat values
Basic Info
- Host: GitHub
- Owner: CellBauhaus
- License: mit
- Language: Julia
- Default Branch: main
- Size: 8.51 MB
Statistics
- Stars: 1
- Watchers: 3
- Forks: 2
- Open Issues: 3
- Releases: 0
Metadata Files
README.md
DLkitty
How to modify:
- Change training/preprocessing functions in
src/execute.jl - Change model structure in
src/neural_net_model.jl
Training and Use
Training
```julia using DLkitty
trainingdf = kcattabletrain() preprocessor = loadpreprocessor() trainedmodel = train(trainingdf, preprocessor; nsamples=1000, nepochs=100) ```
Use
julia
using Statistics
datum = (;
SubstrateSMILES = ["C[C@]12CC[C@H]3[C@H]([C@@H]1CC[C@@H]2O)CCC4=C3C=CC(=C4)O"],
ProteinSequences = ["MAAVKASTSKATRPWYSHPVYARYWQHYHQAMAWMQSHHNAYRKAVESCFNLPWYLPSALLPQSSYDNEAAYPQSFYDHHVAWQDYPCSSSHFRRSGQHPRYSSRIQASTKEDQALSKEEEMETESDAEVECDLSNMEITEELRQYFAETERHREERRRQQQLDAERLDSYVNADHDLYCNTRRSVEAPTERPGERRQAEMKRLYGDSAAKIQAMEAAVQLSFDKHCDRKQPKYWPVIPLKF"],
Temperature = 300.0,
pH = 7.5
)
@show dist = predict_kcat_dist(trained_model, preprocessor, datum)
@show expected = mean(dist) # useful for kinetic modelling
@show upper_bound = quantile(dist, 0.99) # useful for EC FBA
Evaluation
(Better evaluation would use kfold-cross validations splitting from kcat_table_train_and_valid)
```julia using Tables using Distributions
evaldf = filter(iscomplete, kcattablevalid()) evaldf.predictedkcatdists = map(Tables.namedtupleiterator(evaldf)) do datum predictkcatdist(trained_model, preprocessor, datum) end
evaldf.loglikelyhoods = loglikelihood.(evaldf.predictedkcatdists, evaldf.Value) @show loglikelyhoodofevalset = sum(evaldf.loglikelyhoods) # an extremely small number
evaldf.aetomode = abs.(mode.(evaldf.predictedkcatdists) .- evaldf.Value) @show meanaetomode = mean(evaldf.aeto_mode) ```
Owner
- Name: CellBauhaus
- Login: CellBauhaus
- Kind: organization
- Website: https://cellbauhaus.com/
- Repositories: 1
- Profile: https://github.com/CellBauhaus
Citation (CITATION.cff)
# Go to https://citation-file-format.github.io/cff-initializer-javascript/#/ to finish this cff-version: 1.2.0 title: DLkitty.jl message: >- If you use this software, please cite it using the metadata from this file. type: software authors:
GitHub Events
Total
- Issues event: 1
- Watch event: 1
- Delete event: 1
- Push event: 8
- Pull request review event: 1
- Pull request event: 6
- Create event: 92
Last Year
- Issues event: 1
- Watch event: 1
- Delete event: 1
- Push event: 8
- Pull request review event: 1
- Pull request event: 6
- Create event: 92
Dependencies
- julia-actions/cache v2 composite
- julia-actions/setup-julia v2 composite
- actions/checkout v4 composite
- julia-actions/cache v2 composite
- julia-actions/setup-julia v2 composite
- actions/checkout v4 composite
- lycheeverse/lychee-action v1 composite
- actions/checkout v4 composite
- codecov/codecov-action v4 composite
- julia-actions/cache v2 composite
- julia-actions/julia-buildpkg v1 composite
- julia-actions/julia-processcoverage v1 composite
- julia-actions/julia-runtest v1 composite
- julia-actions/setup-julia v2 composite
- JuliaRegistries/TagBot v1 composite