Each gene is preceded by a genetic "marker" called promoter DNA which points to the exact base pair from which to begin transcription. The image from a computerized model shows the double helix structure of promoter DNA with polymerase (pink and blue ribbons) bound to it. Image courtesy of Kim Rasmussen, Condensed Matter and Statistical Physics (T-11)
Laboratory researchers collaborating with colleagues at Harvard Medical School have developed a model and diagnostic tools to simulate the dynamics of DNA. The work is an important step towards beginning to decipher the genetic information contained in the human genome, and could be a significant leap in our understanding of the fundamental processes of life.
The work, led by Los Alamos scientist Kim Rasmussen, has demonstrated the predictive capabilities of the modeling technique through direct comparison with experiments performed on several viral and bacterial DNA sequences by researchers working at Harvard. DNA is deoxyribonucleic acid, the molecules that carry the genetic information essential for the organization and functioning of living cells.
According to Rasmussen, "At a time when the human genome is being sequenced at an amazing rate, it is actually quite remarkable how limited our understanding is of the tremendous amount of information stored in the genome. In other words, we have many of the genetic sequences, but we are largely ignorant of how to decipher them. Our work could be a step toward doing just that."
Genomic sequences consist of only four distinct nucleotides: adenine (A), guanine (G), cytosine (C) and thymine (T). This sequence is strung together by a sugar-phosphate backbone, and stabilized by a complementary strand of DNA that protects each base in the sequence as a pair, wrapped inside the familiar double helix. Genes are the stretches of DNA that contain the blueprints for specific protein and range in length from a few hundred to several thousands of base pairs. Think of them as discrete, linear "files" stored within the genome. The sequential reading of these files by a protein complex is known as transcription. The modeling technique is capable of predicting transcription initiation sites in DNA sequences and may predict the binding sites of important proteins in the transcriptional process.
In addition to Rasmussen, of Condensed Matter and Statistical Physics (T-11), George Kalosakas of T-11 and Alan Bishop of Theoretical (T) Division, along with Chu Hwan Choi and Anny Usheva-Simidjiyska from the Harvard Medical School's Division of Endocrinology developed the modeling technique. Funding for the DNA transcription modeling project was provided by Laboratory-Directed Research and Development (LDRD) funds. LDRD funds basic and applied research and development focusing on creative concepts selected at the discretion of the Laboratory director.