DL-batch-correction

A deep learning model for removing batch effects in microarray data.

Requirements

• matplotlib == 3.4.2
• numpy == 1.20.3
• pandas == 1.2.4
• torch == 1.8.1

Usage

Input

The data is assumed to be in matrix form with columns as samples and features as rows, and \t delimiters.

Expression matrix:

feature   Sample_1 Sample_2 ... Sample_n
Feature_1 -0.1661  -1.2160  ... -0.4630
Feature_2 -1.7001  -0.06567 ...  0.8384
...
Feature_m -0.4630   0.8384  ... -0.8774

Sample labels:

sample   group batch
Sample_1 1     2
Sample_2 1     1
Sample_3 1     2
...
Sample_n 2     2

Notebooks

• NormAE demo
• Peptide transformer
• Transformer Simple Correction (NEW) - Data for use in the notebook can be found within the shared '\MSSHARE\Camilo_Posso\Batch_Correction' folder.