Пожалуй с классической ставшей статьи Molecular ClassiÞcation of
Cancer: Class Discovery and Class Prediction by Gene Expression Monitoring
T. R. Golub,1,2*† D. K. Slonim,1† P. Tamayo,1 C. Huard,1
M. Gaasenbeek,1 J. P. Mesirov,1 H. Coller,1 M. L. Loh,2
J. R. Downing,3 M. A. Caligiuri,4 C. D. BloomÞeld,4
E. S. Lander1,5* - цитируемая в 6799 статьях началась эра классификации злокачественных образований на уровне зкспрессии генов ("It is not enough to say someone has breast cancer or lung cancer, for example. There is a real need to subclassify patients, so that we can match patients very closely with targeted drugs and conduct smarter clinical trials." ) . Авторы в 15 OCTOBER 1999 SCIENCE VOL 286 отметили, что Although cancer classiÞcation has improved over the past 30 years, there has been no general approach for identifying new cancer classes (class discovery) or for assigning tumors to known classes (class prediction). Here, a generic approach to cancer classiÞcation based on gene expression monitoring by DNA microarrays is described and applied to human acute leukemias as a test case.
Our initial leukemia data set consisted of 38 bone marrow samples (27 ALL, 11 AML) obtained from acute leukemia patients at the time of diagnosis...The 6817 genes were sorted by their degree of correlation (16)... The choice to use 50 informative genes in the predictor was somewhat arbitrary....We had expected that the genes most useful in AML-ALL class prediction would simply
be markers of hematopoietic lineage, and would not necessarily be related to cancer pathogenesis. However,...The single most highly correlated gene out of the 6817 genes was the homeobox gene HOXA9, which was overexpressed in patients with treatment failure.
Прекрасная с биологической точки зрения первопроходческая работа оказалась слабоватой ввиду неприменения нужных методов многомерного дискриминантного анализа и дождалась критики специалиста - The oftenly quoted number of 50 genes for classification [Golub, et al. 1999] has no theoretical basis. The number needed depends!
Несмотря на это - награжденная Golub's group has made important discoveries on the molecular basis of several types of leukemia, some of which have resulted in genetic tests that are now standard at most major medical centers worldwide. His team was among the first to use microarrays (also known as DNA chips) for the classification of cancer. "The goal of our work is to develop a new molecular taxonomy of cancer," said Golub. Все таки работу следует считать не завершенной ввиду отсутствия четкой модели ALL, AML с переходом от эмпирического в рациональный уровень.
Та же проблема многомерности -
выделения существенной информации из тотального генома остается и в мировом проекте The Cancer Genome Project - The Cancer Genome Project в котором лаборатории из 11 стран измеряют профиль активности всего генома в 500 образцах различных злокачественных образований.
Чего можно ожидать от проекта и как следует подходить у проблеме?
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
#ASH09 Report: Acute Myeloid Leukemia
Acute myeloid leukemia (AML) is a truly nasty disease and one I hope never to have the misfortune to be diagnosed with.
Last month, at the Chemotherapy Foundation in New York, Dr Norman Wolmark gave an entertaining lecture on what he called the "Decade of Discontent" in colorectal cancer, a bleak period where no new therapies or stunning ideas emerged and the researchers bogged themselves in answering minutiae rather than focusing on the bigger picture.
He could well have been describing AML.
Why?
Take a look at this slide that was shown at the AML Super Friday educational symposium (the reference is from Dohner et al., (2009), published in Blood and how complex the disease has become with a myriad of phenotypes being described:
Picture 56 But this approach begs a most important question:
Which of these mutations or phenotypes are actually relevant and what is driving the cancer's survival and ability to outwit treatment?
No one really knows and thus it illustrates the frustration inherent in making a nasty disease ever more complex. Just because a mutation exists or a pathway is overexpressed does not mean that it is critical to the survival of the leukemia cells! Sometimes the mutations occur as artifacts, a function of a generally increasing leukemic burden over time as the disease gets more established.



