from conftest import output_dir
This repository contains python packages for the AbExp project.
This package contains utility functions for the AbExp project. It contains functions to manipulates polars and spark dataframes, as well as scikit-learn models used by AbExp.
The package can be installed using pip.
# spark
pip install packages/abexp_utils[spark]
# polars
pip install packages/abexp_utils[polars]
# both
pip install packages/abexp_utils[all]Variant effect prediction of promoter variants using the Enformer model.
The package can be installed using pip.
# without cuda support
pip install packages/kipoi_enformer
# with cuda support
pip install packages/kipoi_enformer[gpu]from kipoi_enformer.dataloader import RefTSSDataloader, VCFTSSDataloader
from kipoi_enformer.enformer import Enformer, EnformerAggregator, EnformerTissueMapper, EnformerVeff
from pathlib import Path
from sklearn import linear_model
# define output dirs
output_dir = Path('output')
(output_dir / 'raw/ref.parquet/chrom=chr22').mkdir(exist_ok=True, parents=True)
(output_dir / 'raw/alt.parquet').mkdir(exist_ok=True, parents=True)
(output_dir / 'aggregated/ref.parquet/chrom=chr22').mkdir(exist_ok=True, parents=True)
(output_dir / 'aggregated/alt.parquet').mkdir(exist_ok=True, parents=True)
(output_dir / 'tissue/ref.parquet/chrom=chr22').mkdir(exist_ok=True, parents=True)
(output_dir / 'tissue/alt.parquet').mkdir(exist_ok=True, parents=True)
# define enformer objects
enformer = Enformer()
enformer_aggregator = EnformerAggregator()
enformer_tissue_mapper = EnformerTissueMapper(tracks_path='assets/enformer_tracks/human_cage_enformer_tracks.yaml')
enformer_veff = EnformerVeff(gtf='example_files/annotation.gtf.gz')
# Reference sequences
# define reference dataloader
ref_dl = RefTSSDataloader(fasta_file='example_files/seq.fa', gtf='example_files/annotation.gtf.gz', chromosome='chr22',
canonical_only=False, protein_coding_only=True)
# run enformer on reference genome
enformer.predict(ref_dl, batch_size=2, filepath=output_dir / 'raw/ref.parquet/chrom=chr22/data.parquet',
num_output_bins=11)
# aggregate reference enformer scores
enformer_aggregator.aggregate(output_dir / 'raw/ref.parquet/chrom=chr22/data.parquet',
output_dir / 'aggregated/ref.parquet/chrom=chr22/data.parquet')
# train tissue mapper using reference genome
enformer_tissue_mapper.train([output_dir / 'aggregated/ref.parquet/chrom=chr22/data.parquet'],
output_path=output_dir / 'tissue_mapper.pkl',
expression_path="example_files/isoform_proportions.tsv",
model=linear_model.ElasticNetCV(cv=2))
# map reference to tissues
enformer_tissue_mapper.predict(output_dir / 'aggregated/ref.parquet/chrom=chr22/data.parquet',
output_dir / 'tissue/ref.parquet/chrom=chr22/data.parquet')
# Alternative sequences
# define alternative dataloader
alt_dl = VCFTSSDataloader(fasta_file='example_files/seq.fa', gtf='example_files/annotation.gtf.gz',
vcf_file='example_files/vcf/chr22_var.vcf.gz', variant_upstream_tss=50,
variant_downstream_tss=200, canonical_only=False, protein_coding_only=True)
# run enformer on alternative genome
enformer.predict(alt_dl, batch_size=2, filepath=output_dir / 'raw/alt.parquet/chr22_var.vcf.gz.parquet',
num_output_bins=11)
# aggregate alternative enformer scores
enformer_aggregator.aggregate(output_dir / 'raw/alt.parquet/chr22_var.vcf.gz.parquet',
output_dir / 'aggregated/alt.parquet/chr22_var.vcf.gz.parquet')
# map alternative to tissues
enformer_tissue_mapper.predict(output_dir / 'aggregated/alt.parquet/chr22_var.vcf.gz.parquet',
output_dir / 'tissue/alt.parquet/chr22_var.vcf.gz.parquet')
# Variant effect prediction
enformer_veff.run(ref_paths=[output_dir / 'tissue/ref.parquet/chrom=chr22/data.parquet'],
alt_path=output_dir / 'tissue/alt.parquet/chr22_var.vcf.gz.parquet',
output_path=output_dir / 'veff.parquet', aggregation_mode='canonical')