/MicrobiomeAnalyst-Univariate-Analysis-heatmaps

A personalised script that takes the output a Marker Data profiling single factor analysis output file and generates heatmaps.

Primary LanguagePythonMIT LicenseMIT

Script Guide - WT versus AD comparison

Date: May 2023

  • Author: Mohammed Salim Dason
  • Supervisors: Prof. Davide Cora, Prof. Francesco Favero
  • Institution: Bioinformatics Unit, Università del Piemonte Orientale (UPO) - CAAD

This script was developed to first take the matrix file (output of a different script I created) in tab delimited format that can be fed into Microbiome Analyst and then from the output of single factor analysis at the species taxonomic level, identify the different species that are statistically significant:

  • Absolute log2FC score > 1
  • FDR score < 0.05

Once the statistically significant species have been identified, the script creates temporary files that contain these statically species (see the TEMP folder).

The next step is to lookup each statically significant specie in the input Matrix file, and extract the columns that apply. For example, if the aim is to construct a heatmap of 2 months WT versus 2 months AD, we will extract only the values for samples belonging to this group as shown below (extract from input matrix file):

Name 2WT-2m_S1 3WT-2m_S1 5WT-2m_S2 8WT-2m_S3 9WT-2m_S4 10WT-2m_S5 2AD-2m_S6 4AD-2m_S2 5AD-2m_S7 8AD-2m_S8 9AD-2m_S9 10AD-2m_S10
s__Akkermansia_muciniphila 0 2884 2822 0 9190 0 76 512772 360726 160650 0 0
s__Parasutterella_excrementihominis 0 884 2822 89 93489 0 76 270 36026 150 10 0
s__GGB31312_SGB44628 112 1321 212 12 34 222 76 512 360726 1260 233 131
s__GGB30466_SGB43544 67 27884 2822 0 0 0 76 772 3072 1650 131 0

Once the significant species have been extracted from the corresponding samples, the script will then create a heatmap using a script written in R. This present script uses a python wrapper (util.py) that wraps around the R script that dynamically generates the different heatmaps for the different groups.

Requirements

Before using this script, make sure you have the following requirements:

  • Python 3.7 or higher
  • Pip package manager
  • Input Matrix file for Microbiome Analyst in tab delimited format (Place in data folder)
  • Results of single factor analysis (best to use same naming structure I used)
  • R scripts (3 in total i.e. two_months_heatmap.R, six_months_heatmap.R, twelve_months_heatmap.R)

How to Use

In the create_heatmap.py file, make the following change:

if __name__ == "__main__":
    x = CreateHeatmap("data/{putMatrixFileNameHere}")
    x.start()

Once this is done, run:

$ python create_heatmap.py

Sample Output

image