mfschmidt / pygest

Python Gene Expression Spatial Toolkit

PyGEST: Python Gene Expression Spatial Toolkit

Brief Description

This project serves first as a wrapper to the Allen Brain Institute's human gene expression data. It tracks whether data have been downloaded to local disk, read into memory, etc. and returns pandas dataframes of requested parts of the dataset. If the data are in memory, a reference to the dataframe is returned very quickly. If not, it may take some time to load from disk, or download from brain-map.org first.

It also performs computations comparing gene expression filtered from ABI with other matrices like functional connectivity.

Usage (python3)

This project is still in planning and initial development. Usage will certainly change before the initial release.

Initialize with a path to BIDS structure.

import pygest as ge
from pygest.reporting import sample_overview

data = ge.Data('/home/mike/ge_data')

Now use any of the functions.

probes = data.probes()
samples = data.samples()
expr = data.expression()

args = {"donor": "H03511009", "hemisphere": "L", "ctx": "all"}
pdf_path = sample_overview(data, args, save_as="/home/mike/report.pdf")

Usage (bash)

Data must be structured appropriately, see nonexistent future documentation. ;)

pygest push --probes richiardi --samples cor --comparator /var/fcfile.df -v

Status

The current state of the project is immature. It is untested, in development, and should not yet be used, other than for testing and experimentation. Code that works today may not tomorrow, and vice-versa.