You are here: ProJects >> BioLogy


My PhD thesis is about genetic netwok simulation. You can download it. The work was done in the Deutsches Krebsforschungszentrum in Heidelberg, Germany, in the department for medical and biological informatics. The official project home page is www.cell-o.org.

Some time ago a paper was published by my colleagues, it appeared in the magazine Nature! Read the NaturePaper press release. Congratulations to the the Cell-O team!

This page contains information about cells, genes, proteins, genetic networks, and everything biological you need to know to understand what the genetic network simulation is about. It is not intended for use by expert biologists (i.e. it presents a comprehensible view at the expense of accuracy).


Every living creature consists of cells. In fact, the cell is the smallest unit said to live (this means have metabolism and reproduce). The smallest organsims consist of only one cell (e.g. bacteria, some algae, sone funghi - yeast). Larger organisms, such as you, dear reader, have many millions of cells. These cells are of different types ("fates"). We can distinguish between muscle cells, nerves, and skin cells, for instance.

The fantastic thing is that the genetic information used to create the cells is the same for each cell. Each cell in a human has two times 23 chromosomes, and they contain exactly the same DNA sequence (the same genes) in each and every cell (there are a few exceptions, but I won't go into details here). But if they have the same information to start from, how do the cells become different?

Genes, proteins, and their regulation

If the genes in each cell are identical, there must be something else to make the cells different. This is the way the information is used. Each gene contains the information to make one type of protein. Whether this protein is really produced, is another matter.

There are activator and repressor molecules that regulate the usage of the gentic information. If a repressor is attached to the DNA, the gene at this position is not available, and the protein will not be produced. The gene is said to be "inactive". These activator and repressor molecules are themselves proteins, and the presence of the repressor depends on the activity of the gene coding for it.

If a protein is present (because its corresponding gene is active), it does not mean automatically that the protein is active. Proteins can be activated and inhibited, as well. For example the repressor protein I talked about may be inactive, meaning that at the moment it can't bind to the DNA. The activation (or inhibition) of protein activity can happen by chemical modification of the protein (for example attachment of a phosphate group to the protein molecule) or by binding of a ligand (a smaller molecule or another protein).

There are special proteins calles kinases that are responsible for attaching phosphate groups to other proteins (phosphorylating them) and thereby modulating their activity. Of course, there are also proteins (phosphatases) whose job is the removal of phosphate groups (dephosphorylation).

Okay. We learned:

  • Cells differ from each other by the way they use genetic information. We say they have different genetic states.
  • Genes themselves do nothing. They're just there and wait for someone to access their information and produce new proteins.
  • Proteins can regulate the way the genetic information is used (we say they can activate or deactivate genes)
  • Proteins can also regulate the activity of other proteins.

After having explained all these complicated things, I'm going to simplify it again. We simply ignore the fact that there are genes and proteins, and we assume that there are only genes, and these genes can be either active (meaning that their information is used, and that the protein produced from the gene is active) or inactive (meaning that the genetic information is not used, or that the protein is produced, but inactive). If in fact a protein regulates another, we say in the model that the gene influences the other. The genetic state of a cell in this model is the collection of all the gene's states in the cell.

Genetic networks

The interactions between the genes, i.e. who modifies whose activity, can be represented as a directed, weighted graph. For the non-computer-scientists: this means a picture with arrows connecting the genes (that's why it's directed, one gene influences another, but this other does not necessarily influence the first one). The interactions may be of different strength (weight), as one gene's influence may outweight another's; if gene A says "I activate gene X", and gene B says "I inhibit gene X", one of the two interactions will be stronger. For example, if A → X is stronger than B -> X, X will be active, even if B does its best to inhibit X.