A 25-Signal Proteomic Signature and Outcome for Patients With Resected Non–Small-Cell Lung Cancer
J. Natl. Cancer Inst. Yanagisawa et al.
99: 858
Supplementary Data
Files in this Data Supplement:
- Supplementary Table 1 -
The top 25-ranked mass spectrometry peaks with the accumulated rank in each cross-validation procedure
- Supplementary Fig. 1 -
Assessment of the outcome classifier in the training cohort using the weighted voting algorithm. Blue circles = patients predicted to be at low risk; red circles = patients predicted to be at high risk with the outcome classifier; circles in the green-shaded box to the left = patients alive at 5 years after surgery; and circles in the gray-shaded box to the right = patients who died within 5 years after surgery.
- Supplementary Fig. 2 -
Average number of misclassifications per permutation in each classifier constructed with a defined number of mass spectrometry (MS) signals. Use of the 25 MS signals gave the minimum number of errors (11.1 per permutation) and thus 89% accuracy for classifying patients as having low-risk or high-risk signatures. Dots represent the number of misclassified patients (y axis) as a factor of each number of proteomic signals (x axis).
- Supplementary Fig. 3 -
Unsupervised hierarchical clustering analysis of data in the test cohort by use of the mass spectrometry signals selected as a discriminator between non-small cell lung cancer (NSCLC) and normal lung (NL) tissues in the training cohort (58 tumor tissue samples and seven normal lung tissue samples). We used top 40 signals for this analysis.
