soedinglab/gxpred

Name: gxpred

Owner: Söding Lab

Description: Development branch of GxPRED

Created: 2017-10-11 11:42:32.0

Updated: 2017-11-22 11:53:25.0

Pushed: 2017-10-30 13:40:36.0

Homepage:

Size: 1720

Language: Python

GitHub Committers

User	Most Recent Commit	# Commits

Other Committers

User	Email	Most Recent Commit	# Commits

README

GxPRED

This is the development branch of Gene Expression Prediction Algorithm, under the production name GxPRED.

What is this repository for?

• Development of GxPRED
• v0.01

How to run?

GxPRED provides core utilities / API for learning and predicting gene expression levels from genotype. Examples are shown in the base directory for learning from gEUVADIS data (learn_from_geuvadis.py) and predicting on GTEx data (predict_on_gtex.py). You can adapt these files for any dataset.

CODEBASE = "/path/to/this/directory"
TRAINVCF = "/path/to/gz/vcf/file/for/training"
TRAINRPKM = "/path/to/normalized/gene/expression/file/for/training"
TRAINGTF = "/path/to/gtf/file"
CHROM = "21" # change it to whichever chromosome you are interested 
PREDVCF = "/path/to/gz/vcf/for/prediction/samples"
MODELDIR = "/path/to/directory/where/model/will/be/saved"

python ${CODEBASE}/learn_from_geuvadis.py   --vcf ${TRAINVCF} --rpkm ${TRAINRPKM} --gtf ${TRAINGTF} --chrom ${CHROM} --params 0.01 0.0 0.01 0.001 0.001 --outdir ${MODELDIR}
python ${CODEBASE}/predict_on_gtex.py --vcf ${PREDVCF}  --model ${MODELDIR} --chrom ${CHROM} --outprefix predicted_gene_expression