The classification model for the lung_SQ_AD data set was built with Adenocarcinoma, Squamous cell carcinoma and normal lung tissue samples from the original dataset (GSE1987). The Adenosquamous sample was added to the Adenocarcinoma class. The metastasis samples were excluded from our analysis, because of their small number (3 samples).
Platform: Affymetrix GeneChip Human Genome U95 Version [1 or 2] Set HG-U95A
Number of genes: 10541 Number of samples: 34 Note: From the originally measured 12625 probe sets we removed genes that were not present (P) in at least one sample
Predictive accuracy with 10-fold cross validation (classifying using the best projection with eight attributes):
Following are the three best-ranked visualization with eight, six and four attributes in respect to the visualization score, that is, visualizations where examples from different diagnostic classes are best separated:
Score: 99.14% Genes: 35454_at: phospholipase C-like 4, PLCL4 36280_at: granzyme K (serine protease, granzyme 3; tryptase II), GZMK 34301_r_at: keratin 17, KRT17 36186_at: RNA binding protein S1, serine-rich domain, RNPS1 34800_at: leucine-rich repeats and immunoglobulin-like domains 1, LRIG1 34928_at: zinc finger protein 205, ZNF205 36939_at: glycoprotein M6A, GPM6A 37983_at: angiotensin II receptor, type 1, AGTR1
Score: 98.06% Genes: 36280_at: granzyme K (serine protease, granzyme 3; tryptase II), GZMK 34301_r_at: keratin 17, KRT17 36244_at: zinc finger protein 239, ZNF239 38582_at: serine protease inhibitor, Kazal type 1, SPINK1 37196_at: cadherin 5, type 2, VE-cadherin (vascular epithelium), CDH5 37983_at: angiotensin II receptor, type 1, AGTR1