FRI > Biolab > Supplements

Data set name: braintumor


Original data set (Pomeroy et al.)
Data set for Orange
Brief description:
We built a classification model that attempts to distinguishes between different embryonal tumors of the central nervous system on the basis of DNA expression signatures. We used the dataset A1 from the original research done by Pomeroy et al., consisting of 40 samples from 5 diagnostic classes (medulloblastomas, malignant gliomas, atypicalteratoid/rhabdoid tumors, primitive neuroectodermal tumors and normal cerebella).

Platform: Affymetrix HuGeneFL array

Diagnostic classes:
- medulloblastoma (medulloblastoma): 10 examples (25.0%)
- malignant glioma (glioma): 10 examples (25.0%)
- Rhabdoid tumor (RhabdoidTu): 10 examples (25.0%)
- normal cerebellum (Normal): 4 examples (10.0%)
- primitive neuroectodermal tumor (PNET): 6 examples (15.0%)
Number of genes: 7129
Number of samples: 40
Predictive accuracy with 10-fold cross validation (classifying using the best projection with eight attributes):
Classification accuracy: 52.50%
Area under curve (AUC): 0.837
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: 95.31%
Genes:
U44060_at: prospero-related homeobox 1, PROX1
M30448_s_at: fibrillarin, FBL
X13916_at: low density lipoprotein-related protein 1 (alpha-2-macroglobulin receptor), LRP1
M93426_at: "protein tyrosine phosphatase, receptor-type, Z polypeptide 1", PTPRZ1
D87463_at: phytanoyl-CoA hydroxylase interacting protein, PHYHIP
M12125_at: tropomyosin 2 (beta), TPM2
J04164_at: interferon induced transmembrane protein 1 (9-27), IFITM1
M80397_s_at: "polymerase (DNA directed), delta 1, catalytic subunit 125kDa", POLD1
Score: 91.26%
Genes:
M93119_at: insulinoma-associated 1, INSM1
D26070_at: "inositol 1,4,5-triphosphate receptor, type 1", ITPR1
X86693_at: "SPARC-like 1 (mast9, hevin)", SPARCL1
X13916_at: low density lipoprotein-related protein 1 (alpha-2-macroglobulin receptor), LRP1
X79683_s_at: "laminin, beta 2 (laminin S)", LAMB2
X14830_at: "cholinergic receptor, nicotinic, beta polypeptide 1 (muscle)", CHRNB1
Score: 82.26%
Genes:
M30448_s_at: fibrillarin, FBL
M93426_at: "protein tyrosine phosphatase, receptor-type, Z polypeptide 1", PTPRZ1
M12125_at: tropomyosin 2 (beta), TPM2
D87463_at: phytanoyl-CoA hydroxylase interacting protein, PHYHIP

Attribute ranking

Following is the histogram of genes showing how often are they present in one of the top 100 radviz visualizations with 8 attributes.

Genes:
M93119_at: insulinoma-associated 1, INSM1
M12125_at: tropomyosin 2 (beta), TPM2
X86693_at: "SPARC-like 1 (mast9, hevin)", SPARCL1
U90902_at: T-cell lymphoma invasion and metastasis 1, TIAM1
U14968_at: ribosomal protein L27a, RPL27A
D26070_at: "inositol 1,4,5-triphosphate receptor, type 1", ITPR1
X63578_rna1_at: parvalbumin, PVALB
M32304_s_at: TIMP metallopeptidase inhibitor 2, TIMP2
U48250_at: oligodendrocyte lineage transcription factor 2, OLIG2
D87463_at: phytanoyl-CoA hydroxylase interacting protein, PHYHIP
X13916_at: low density lipoprotein-related protein 1 (alpha-2-macroglobulin receptor), LRP1
U44060_at: prospero-related homeobox 1, PROX1
L35592_at: Similar to expressed sequence AW125688, ---
U40372_at: "phosphodiesterase 1C, calmodulin-dependent 70kDa", PDE1C
L38969_at: thrombospondin 3, THBS3
M93426_at: "protein tyrosine phosphatase, receptor-type, Z polypeptide 1", PTPRZ1
X79683_s_at: "laminin, beta 2 (laminin S)", LAMB2
Z15108_at: "protein kinase C, zeta", PRKCZ
L33243_at: polycystic kidney disease 1 (autosomal dominant), PKD1
J04469_at: "creatine kinase, mitochondrial 1B", CKMT1B