Installation#

Install from source#

git clone https://github.com/algaebrown/Metadensity.git
cd Metadensity
# Install dependencies
conda env create -n Metadensity --file environment.yaml
conda activate Metadensity
# Install Metadensity
cd Metadensity
pip install -e .

Build annoatations#

  1. Step 1: Download annoatations

# on bash, download gencode
wget https://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_40/gencode.v40.primary_assembly.annotation.gff3.gz
wget https://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_40/GRCh38.primary_assembly.genome.fa.gz
# Unzip it
gunzip gencode.v40.primary_assembly.annotation.gff3.gz

2. compile config.ini file (Using gencode 40 as an example) Example file is in here

This is a file that stores all the annoatations

[FILES]
GENOME_FA=/home/hsher/gencode_coords/GRCh38.p13.genome.fa
GENCODE=/home/hsher/gencode_coords/gencode.v33.annotation.gff3
# OTHER
BRANCHPOINT=/home/hsher/gencode_coords/branchpoint_hg38.bed
BRANCHPOINT_PRED=/home/hsher/gencode_coords/gencode_v26_branchpoints_pred.csv
POLYA=/home/hsher/gencode_coords/polyA.atlas.clusters.2.0.GRCh38.96.bed
MIRNA=/home/hsher/gencode_coords/miR/hsa_hg38.gff3
SNORNA=/home/hsher/gencode_coords/snoDB_hg38.tsv
LNCRNA=/home/hsher/gencode_coords/lncipedia_5_2_hc_hg38.gff
# processed from GENCODE
TRANSCRIPT=/home/hsher/gencode_coords/gencode.v33.transcript.gff3
FEATURE=/home/hsher/gencode_coords/gencode.v33.combine.sorted.gff3
# PARSED PICKLE
DATADIR=/home/hsher/projects/Metadensity/metadensity/data/hg38

  • Fill in GENOME_FA, GENCODE as the files that you just downloaded.

  • The fields under OTHER are optional. See here for download link

  • TRANSCRIPT and FEATURE is generated from the following steps. Simply replace .gff with .transcript.gff and combined.sorted.gff3

  • Lastly, DATADIR is also generated from the steps below. Simply put any directory that can hold lots of data.

3. Generate DATADIR use this notebook to generate. This notebooks takes CONFIG.ini and parse all the annotations into a dictionary.

4. Generate TRANSCRIPT and FEATURE from GENCODE and UCSC canonical transcripts Download canonical transcripts here. Use this script to generate TRANSCRIPT and FEATURE files. .. code-block:

module load bedtools
bash gencode_canon_filtering.sh gencode.v40.primary_assembly.annotation.gff3 knownCanonical.txt

Run the test notebook#

Run this test notebook.