the doctor is. Microarrays may become increasingly important for the diagnosis of cancer.
to choose the best therapy for spreading the cancer, doctors need to know which created a fabric. Now researchers trained a computer program to distinguish DNA of four seemingly similar deadly tumors. The method could lead to a faster and more appropriate treatment diagnosis for a variety of cancers.
PinPoint pathologists often the origin of cancer by studying the size, shape and color of the cells under a microscope. This is a delicate task for any cancer cells that are alike, as a group of four rare childhood cancers collectively called tumors with small round blue cells. In these cases, doctors often use several spots that bind diagnostic proteins, which can delay diagnosis. Betting that the genetic fingerprint would be faster and more accurate molecular biologist Paul Meltzer and pediatric oncologist Javed Khan of the National Human Genome Research Institute (NHGRI) in Bethesda, Maryland, and colleagues examined gene activity in four cancers - neuroblastoma, rhabdomyosarcoma, non-Hodgkin lymphoma, and Ewing tumor family.
The team interviewed 6567 genes at once using DNA chips. On these chips called genes, the activity of each gene is represented by a colored dye intensity. To determine the expression profile of genes, the researchers then ran the color intensity data through a so-called artificial neural network, a program modeled on the behavior of neurons in the brain that can learn by trial and error. "You get a complete picture of what is happening with this cancer," says Meltzer.
After the program has learned the characteristics of each type of cancer, it has correctly identified all 20 test samples . He also correctly rejected five samples of other types of cancer as well as healthy cells, reports the team in the June issue of Nature Medicine .
"I think it's very encouraging, "said Lee Hartwell cell biologist research on cancer Fred Hutchinson in Seattle. Although the sample size is small, he warns, the work clearly shows that cancers the same cell and tissue type are similar enough to distinguish them from other cancers, despite genetic differences between people, he said. "If it continues to support these microarrays will be very, very important for the diagnosis, "he said.
Related Sites
project Overview NHGRI microarray
homepage Paul Meltzer
An introduction to neural networks
0 Komentar