Compute differentially expressed genes using Hypergeometric test (Affymetrix probes) (workflow)

Workflow title

Compute differentially expressed genes using Hypergeometric test (Affymetrix probes)

Provider

GeneXplain GmbH

Workflow overview

Description

This workflow is designed to identify upregulated and downregulated genes for experimental data with any number of data points for each experiment and control.  It can be used even for the cases with one data point in each experiment and control.

As input, the normalized data with Affymetrix probeset IDs can be submitted.

Such normalized files are resulting from the output of the “Normalize data” procedure under  “analyses/Methods/Data normalization/Normalize Affymetrix experiment and control”.

At the next step, p-value is calculated for up-and down-regulated probeset IDs. This workflow applies hypergeometric analysis for p-value calculation.

Simultaneously, log fold change is calculated for each probeset ID, and as the result of this step, a table is produced in which both LogFoldChange and p-value are assigned to each probeset ID.

Further, this table is filtered by several conditions in parallel, to identify upregulated, downregulated, as well as a joint table of up- & downregulated Affymetrix probeset IDs.

The filtering criteria are set as the following.

For upregulated probes: LogFoldChange>0.5 and -log_P_value_>3.

For downregulated probes: LogFoldChange<-0.5 and -log_P_value_<-3.

For up- & downregulated probes: (LogFoldChange>0.5 and -log_P_value_>3 & LogFoldChange<-0.5 and -log_P_value_<-3)

Resulting tables of the upregulated, downregulated, and up- & downregulated Affymetrix probeset IDs are annotated with additional information, gene description, gene symbols, species.

Finally, these tables are converted into the tables of genes. Two tables are produced, with Ensembl Gene IDs and with Entrez IDs.

Parameters

Experiment normalized

Control normalized

Species

Results folder