Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2001 Dec 18;98(26):15149-54.
doi: 10.1073/pnas.211566398. Epub 2001 Dec 11.

Multiclass cancer diagnosis using tumor gene expression signatures

S Ramaswamy  1 , P Tamayo, R Rifkin, S Mukherjee, C H Yeang, M Angelo, C Ladd, M Reich, E Latulippe, J P Mesirov, T Poggio, W Gerald, M Loda, E S Lander, T R Golub
Affiliations

Multiclass cancer diagnosis using tumor gene expression signatures

S Ramaswamy et al. Proc Natl Acad Sci U S A. .

Abstract

The optimal treatment of patients with cancer depends on establishing accurate diagnoses by using a complex combination of clinical and histopathological data. In some instances, this task is difficult or impossible because of atypical clinical presentation or histopathology. To determine whether the diagnosis of multiple common adult malignancies could be achieved purely by molecular classification, we subjected 218 tumor samples, spanning 14 common tumor types, and 90 normal tissue samples to oligonucleotide microarray gene expression analysis. The expression levels of 16,063 genes and expressed sequence tags were used to evaluate the accuracy of a multiclass classifier based on a support vector machine algorithm. Overall classification accuracy was 78%, far exceeding the accuracy of random classification (9%). Poorly differentiated cancers resulted in low-confidence predictions and could not be accurately classified according to their tissue of origin, indicating that they are molecularly distinct entities with dramatically different gene expression patterns compared with their well differentiated counterparts. Taken together, these results demonstrate the feasibility of accurate, multiclass molecular cancer classification and suggest a strategy for future clinical implementation of molecular cancer diagnostics.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Clustering of tumor gene expression data and identification of tumor-specific molecular markers. Hierarchical clustering (a) and a 5 × 5 self-organizing map (SOM) (b) were used to cluster 144 tumors spanning 14 tumor classes according to their gene expression patterns. (c) Gene expression values for class-specific OVA markers, as determined using the S2N metric, are shown. Columns represent 190 primary human tumor samples ordered by class. Rows represent 10 genes most highly correlated with each OVA distinction. Red indicates high relative level of expression, and blue represents low relative level of expression. The known cancer markers prostate-specific antigen (PSA), carcinoembryonic antigen (CEA), and estrogen receptor (ER) are identified. BR, breast adenocarcinoma; PR, prostate adenocarcinoma; LU, lung adenocarcinoma; CR, colorectal adenocarcinoma; LY, lymphoma; BL, bladder transitional cell carcinoma; ML, melanoma; UT, uterine adenocarcinoma; LE, leukemia; RE, renal cell carcinoma; PA, pancreatic adenocarcinoma; OV, ovarian adenocarcinoma; ME, pleural mesothelioma; CNS, central nervous system.
Figure 2
Figure 2
Multiclass classification scheme. The multiclass cancer classification problem is divided into a series of 14 OVA problems, and each OVA problem is addressed by a different class-specific classifier (e.g., “breast cancer” vs. “not breast cancer”). Each classifier uses the SVM algorithm to define a hyperplane that best separates training samples into two classes. In the example shown, a test sample is sequentially presented to each of 14 OVA classifiers and is predicted to be breast cancer, based on the breast OVA classifier having the highest confidence.
Figure 3
Figure 3
Multiclass classification results. (a) Results of multiclass classification by using cross-validation on a training set (144 primary tumors) and independent testing with 2 test sets: Test (54 tumors; 46 primary and 8 metastatic) and PD (20 poorly differentiated tumors; 14 primary and 6 metastatic). (b) Scatter plot showing SVM OVA classifier confidence as a function of correct calls (blue) or errors (red) for Training, Test, and PD samples. A, accuracy of prediction; %, percentage of total sample number.
Figure 4
Figure 4
Multiclass classification error analysis. Matrices delineate distribution of actual compared with predicted class membership for multiclass prediction on training (crossvalidation) and test sets.
Figure 5
Figure 5
Multiclass classification as a function of gene number. Training and test datasets were combined (190 tumors; 14 classes), then were randomly split into 100 training and test sets of 144 and 46 samples (all primary tumors) in a class-proportional manner. SVM OVA prediction was performed, and mean classification accuracy for the 100 splits was plotted as a function of number of genes used by each of the 14 OVA classifiers, showing decreasing prediction accuracy with decreasing gene number. Results using other algorithms (k-NN, k-nearest neighbors; WV, weighted voting) and classification schemes (AP, all-pairs) are also shown.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Ramaswamy S, Osteen R T, Shulman L N. In: Clinical Oncology. Lenhard R E, Osteen R T, Gansler T, editors. Atlanta: Am. Cancer Soc.; 2001. pp. 711–719.
    1. Tomaszewski J E, LiVolsi V A. Cancer. 1999;86:2198–2200. - PubMed
    1. Connolly J L, Schnitt S J, Wang H H, Dvorak A M, Dvorak H F. In: Cancer Medicine. Holland J F, Frei E, Bast R C, Kufe D W, Morton D L, Weichselbaum R R, editors. Baltimore: Williams & Wilkins; 1997. pp. 533–555.
    1. Golub T R, Slonim D K, Tamayo P, Huard C, Gaasenbeek M, Mesirov J P, Coller H, Loh M L, Downing J R, Caligiuri M A, et al. Science. 1999;286:531–537. - PubMed
    1. Alizadeh A A, Eisen M B, Davis R E, Ma C, Lossos I S, Rosenwald A, Boldrick J C, Sabet H, Tran T, Yu X, et al. Nature (London) 2000;403:503–511. - PubMed

Publication types

Substances