https://github.com/astrazeneca/gim
gene interaction matrices, a novel approach to using ConvNets on gene expression data
https://github.com/astrazeneca/gim
Last synced: 2 months ago
JSON representation
gene interaction matrices, a novel approach to using ConvNets on gene expression data
- Host: GitHub
- URL: https://github.com/astrazeneca/gim
- Owner: AstraZeneca
- License: apache-2.0
- Created: 2021-08-27T08:31:26.000Z (almost 5 years ago)
- Default Branch: master
- Last Pushed: 2022-06-22T04:41:08.000Z (almost 4 years ago)
- Last Synced: 2025-01-18T23:00:07.348Z (over 1 year ago)
- Language: Jupyter Notebook
- Size: 95.7 KB
- Stars: 1
- Watchers: 3
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
Awesome Lists containing this project
README
# GIM
# Introduction
We developed Gene Interaction Matrices (GIM) as a biologically inspired, gene-interaction based data transformation on gene expression data to create an image-like feature matrix from any gene expression-based study. The transformed data can then be used with any CNN based machine learning approaches for a variety of challenging problems such as disease diagnostics and drug development.
# Dependencies
```
python >= 3.7
numpy 1.20.3
pandas 1.3.0
```
# Usage
An example workflow to create a GIM is described here. To begin import the necessary dependencies and gim .py file containing the transform function.
```
import pandas
import gim
```
Load the treatment and control files containing one or more replicates from the gim/data/ directory.
```
df_control_replicates = pd.read_csv("control_replicates.csv")
df_treatment_replicates = pd.read_csv("treatment_replicates.csv")
```
Apply the gim_transform function to the files.
```
sample_img = gim.gim_transform(df_control_replicates, df_treatement_replicates)
```
# Reference
Pre-print: https://www.biorxiv.org/content/10.1101/2021.09.07.459284v1