Skip to content
Snippets Groups Projects
Select Git revision
  • 1.7.0
  • 1.7.0b
  • 1.7.0c
  • 2017-11-13
  • 20180114
  • 201911
  • 20200121_pre_presentation_refactor
  • 202010-phospho
  • 3-flank-variant
  • add-class1-ensemble
  • add-class1-ensemble-rebased
  • add-hyperparameters-base
  • antigen-presentation-updates
  • arch2
  • cartesian-product-cli
  • dataset-object
  • docs-pan-motifs
  • fix-commandline-predict-peptide-lengths
  • gh-pages
  • improved-allele-parsing
  • pre-2.1
  • v2.0.1
  • v2.0.0
  • pre-2.0
  • 1.6.1
  • pre-1.7.0
  • pre_presentation_refactor
  • 1.6.0
  • 1.4.0
  • pre-1.4.0
  • pan-dev1
  • 1.2.2
  • 1.2.0
  • pre-1.2.1
  • pre-1.2
  • pre-1.1
  • 1.0.0
  • pre-1.0
  • 0.9.2
  • 0.9.1
40 results
You can move around the graph by using the arrow keys.
Created with Raphaël 2.2.01Dec30Nov28272625241514139813Oct16Aug1510331Jul3027241926Jun2213121125May2423222322211817151017Apr24Mar232221201917161411632128Feb262319161514131211109849Jan12Dec115Oct26Sep171615130Aug3118Jul1224Jun231732121May1918171413111096543229Apr2821201918718Feb171617Dec1313Nov124330Oct2826232221201917151412643216Sep28Jul1413bump mhcflurry versionpre-1.0pre-1.0fix requirementslazy load kerasBump to Keras==2.1.2Fix encoding for 0-length sequencesMore speed efficient variable-length peptide encodingSwitch back to dense layer size 32 from 16fixupdate speed benchmarkfixAdd time remaining estimation to training scriptupdate models_class1_experiments1 to use minibatch_size=128Free up memory: get rid of train data as models are trainedbetter progress printingPrint progress no more often than once every 5 seconds while trainingClearer output during trainingExperimental support for parallelization (Theano only) in train_allele_specific_models_commandimplement Class1AffinityPredictor.merge() methodbump pandas requirementUse 1M peptides for calibrationarg tweakAdd minibatch_size hyperparameterChange prediction batch_size from 32 to 4096 for better performancepeptide encoding optimizationMuch faster performance (~5X improvement) in amino acid encodingfixmore efficient peptide encoding in calibrate_percentile_ranksupdateSwitch from 32 hidden layer size to 16 in standard modelsAdd cross-validation GENERATE.sh and related scriptsadd widelocal variantfixupdate models_class1_experiments1 to include embedding and onehot input variationsupdate models_class1_experiments1json -> yaml renamesnew architecturesFix tests, add logging to calibrate_percentile_ranks, add SIGUSR1 trick to train commandREADME update (mhctools mention) and percentile rank tweaksfixAdd throw argument to percentile_ranks