Implement UTR dataloaders and generalized protein dataloader

v0.4.0

Bump version: 0.3.5 → 0.4.0

Last version supporting python 3.5

v0.3.5

Bump version: 0.3.4 → 0.3.5

v0.2.3

• removed some dependencies (hdf5 and related )
• inital api for vcf_seq_extractor added
• minor changes related to circleci
• usage of kipoi_conda and kioi_utils packages

v0.2.2

Features

• added vcf_seq extractors #23
  • SingleSeqVCFSeqExtractor
  • SingleVariantVCFSeqExtractor
  • VariantSeqExtractor

Refactoring

• kipoiseq now uses kipoi-utils and kipoi-conda #22

v0.2.1

Refactoring

• rename IntervalSeqDl -> SeqIntervalDl and IntervalSeqStringDl -> StringSeqIntervalDl

v0.2.0

Refactoring

• Rename dataloaders:
  • closes #13
  • datasets.py -> dataloaders.py
  • SeqDataset -> IntervalSeqDl
  • SpliceDataset -> MMSpliceDl
• remove use_strand

New features

• MMSpliceDl - MMSplice dataloader
• add get_output_shape
• Allow to use genomelake's one_hot_encoding when genomelake is installed

v0.1.1

• enable dataset dataloaders to be used with variant effect prediction.