I am building a project on Inferring Gene Regulatory Networks using ARACNE and PCA-CMI algorithms, and the input to these algorithms is taken from the DREAM3 challenge.

The format of the input data is shown in the image.

Now according to what I studied, a Gene Expression Matrix has its rows represent genes, columns represent samples such as tissues or experimental conditions and the numbers in each cell refer to the expression level of a particular gene in the particular sample.

And that Gene Expression is the process in which information from a gene is synthesized to obtain Gene Products. The process of Gene Expression is that it undergoes a transcription process where a transcription factor attaches itself to the gene and then results in the formation of gene product.

But the input data apparently shows gene-gene interactions.

As a result of this I am extremely confused.

Input Data format of DREAM3Any kind of help would be greatly appreciated. Thanks.


2 Answers 2


I don't know anything about your algorithms. But i will try to explain your the format of data that was given to you. I don't know your background in Biology, so i will assume it is not your expertise field and will make some simplifications of the subject.

What is gene expression?
To understand what is a gene expression, you need to understand what is a gene:

A gene is a sequence in the DNA (composed of 4 bases A,T,C and G) that can be transcribed by a protein, in your context we will say it always start with a start codon (a codon is a triplet of DNA bases) and stop with a stop codon. It is usually around a thousand bases long. The transcription will give you a RNA, and that RNA can be translated (note the difference with transcribed) into a whole new protein.

Now the gene Expression is a measure of the amount of RNA from the gene you are looking for. In a cell you can have around 10000-100000 copies of that RNA; the raw count is not really stable as you can have extract two cells or three and it will change your "expression". Most of the time we normalize the count by the count of a bunch of known genes called housekeeping genes. The particularity of these genes is that their expression is quite stable.

The figures you have are a ratio between RNA copies of your gene of interest and RNA copies of a stable gene (in term of expression).

What does G1(-/-) and wt mean?

Usually when we (biologist) call a sample wt it means wild type, in other words, a normal individual. G1(-/-) means that this sample is from an individual with the G1 gene inactive (could be obtained by genetic modification).

What is a gene-gene interaction?
An example of interaction first: let's imagine a gene coding for an enhancer (booster) of transcription. the presence of this gene allow another gene to be expressed in the cell, if this gene isn't present or isn't transcribed, the other gene won't be expressed either.

How can we see interaction there?

Let's look at line one, the control (normal individual), We see that G4 (column) is almost unexpressed (0.07). In normal conditions, this gene is not expressed. Now let's look at mutant G1(-/-) and G6(-/-), we see the expression of G4 is rising (0.37;0.23) compared to the control (wt). We can suppose than G1 and G6 are acting as inhibitors of the G4 gene.


The short answer is yes, of course there are gene-gene interactions. I'm not sure what the point of the experiment would be if there were not.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .