🧬 Integration of multiple gene expression with GeneCompete 🏆

GeneCompete is a tool to combine heterogeneous gene expression datasets to order gene importance.

Web-Application

The easy-to-use web-based platform can be accessed through 👉 https://genecompete.streamlit.app/ 👈

Note

We suggest using Python function for large datasets.

Python functions

We proposed two python function. The output of these functions are pandas dataframe with rating and ranking scores of each gene.

• GeneCompete_Union ¹
• GeneCompete_Intersect ²

Input	Description
table	Gene expression data: Multiple files where the first column is gene name. These data can be prepared by any tools.
name	Column name: The interested value that will be used as competing score (in the example is logFC).
method	Ranking Method: Select 'Win-loss', 'Massey', 'Colley', 'Keener', 'Elo', 'Markov', 'PageRank', or 'BiPagerank'
reg	Regulation cases: 'Up-regulation' or 'Down-regulation'
FC	logFC threshold: The large number of genes can consume computational time. Before ranking, datasets are filtered with logFC > FC in case of up-regulation and logFC < -FC for down-regulation.

• Installation !git clone https://github.com/panisajan/GeneCompete
• Load data

import pandas as pd
dat1 = pd.read_csv("sample_data/dat1.csv", index_col=0)
dat2 = pd.read_csv("sample_data/dat2.csv", index_col=0)
dat3 = pd.read_csv("sample_data/dat3.csv", index_col=0)
dat4 = pd.read_csv("sample_data/dat4.csv", index_col=0)
dat1.head()

Example 1 (Union strategy):

from GeneCompete_Union import*

my_data = [dat1, dat3, dat4]
my_methods = ['Win-loss', 'Massey', 'BiPagerank']

score = GeneCompete_Union(table = my_data, name = 'logFC', method = my_methods, reg = 'Down-regulation', FC = 1)
score.head()

Example 2 (Intersect strategy):

from GeneCompete_Intersect import*

my_data = [dat1, dat2, dat3, dat4]
my_methods = ['Win-loss', 'Keener', 'PageRank']

score = GeneCompete_Intersect(table = my_data, name = 'logFC', method = my_methods, reg = 'Up-regulation', FC = None)
score.head()

Further Reading

Footnotes

1. aa ↩

2. aa ↩