Data set name: MLL


Original data set (Armstrong et al.)
Data set for Orange
Brief description:
Mixed-lineage leukemias (MLL) are a subset of human acute lymphoblastic leukemias with a chromosomal translocation involving the mixed-lineage leukemia gene. MLL translocations are typically found in infant leukemias and in chemotherapy-induced leukemias and have a particularly poor prognosis. The original research on this dataset (Armstrong et al.) suggested, that MLL have a highly uniform and distinct pattern that clearly distinguishes them from conventional acute lymphoblastic (ALL) or acute myeloid leukemias (AML). Our classification model also shows clear separation of the three diagnostic classes (ALL, AML and MLL leukemias) based on gene expression values.

Platform: Affymetrix Human Genome U95Av2 Array

Diagnostic classes:
- acute lymphoblastic leukemia (ALL): 24 examples (33.3%)
- mixed-lineage leukemia (MLL): 20 examples (27.8%)
- acute myeloid leukemia (AML): 28 examples (38.9%)
Number of genes: 12533
Number of samples: 72
Predictive accuracy with 10-fold cross validation (classifying using the best projection with eight attributes):
Classification accuracy: 91.61%
Area under curve (AUC): 0.973
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: 99.89%
Genes:
1389_at: "membrane metallo-endopeptidase (neutral endopeptidase, enkephalinase, CALLA, CD10)", MME
39556_at: "spectrin, beta, non-erythrocytic 1", SPTBN1
963_at: "ligase IV, DNA, ATP-dependent", LIG4
1065_at: fms-related tyrosine kinase 3, FLT3
1914_at: cyclin A1, CCNA1
39598_at: "gap junction protein, beta 1, 32kDa (connexin 32, Charcot-Marie-Tooth neuropathy, X-linked)", GJB1
239_at: cathepsin D (lysosomal aspartyl peptidase), CTSD
1894_f_at: ---, ---
Score: 99.64%
Genes:
1389_at: "membrane metallo-endopeptidase (neutral endopeptidase, enkephalinase, CALLA, CD10)", MME
963_at: "ligase IV, DNA, ATP-dependent", LIG4
1065_at: fms-related tyrosine kinase 3, FLT3
36777_at: "killer cell lectin-like receptor subfamily K, member 1", KLRK1
39598_at: "gap junction protein, beta 1, 32kDa (connexin 32, Charcot-Marie-Tooth neuropathy, X-linked)", GJB1
239_at: cathepsin D (lysosomal aspartyl peptidase), CTSD
Score: 98.83%
Genes:
1389_at: "membrane metallo-endopeptidase (neutral endopeptidase, enkephalinase, CALLA, CD10)", MME
963_at: "ligase IV, DNA, ATP-dependent", LIG4
34306_at: muscleblind-like (Drosophila), MBNL1
39448_r_at: B7 gene, B7

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:
1389_at: "membrane metallo-endopeptidase (neutral endopeptidase, enkephalinase, CALLA, CD10)", MME
34306_at: muscleblind-like (Drosophila), MBNL1
36162_at: basigin (OK blood group), BSG
963_at: "ligase IV, DNA, ATP-dependent", LIG4
39598_at: "gap junction protein, beta 1, 32kDa (connexin 32, Charcot-Marie-Tooth neuropathy, X-linked)", GJB1
40797_at: ADAM metallopeptidase domain 10, ADAM10
39448_r_at: B7 gene, B7
1894_f_at: ---, ---
39931_at: dual-specificity tyrosine-(Y)-phosphorylation regulated kinase 3, DYRK3
32847_at: "myosin, light polypeptide kinase", MYLK
36239_at: "POU domain, class 2, associating factor 1", POU2AF1
39556_at: "spectrin, beta, non-erythrocytic 1", SPTBN1
34168_at: "deoxynucleotidyltransferase, terminal", DNTT
40570_at: forkhead box O1A (rhabdomyosarcoma), FOXO1A
39385_at: "alanyl (membrane) aminopeptidase (aminopeptidase N, aminopeptidase M, microsomal aminopeptidase, CD13, p150)", ANPEP
266_s_at: CD24 antigen (small cell lung carcinoma cluster 4 antigen), CD24
39011_at: endosulfine alpha, ENSA
1065_at: fms-related tyrosine kinase 3, FLT3
39318_at: T-cell leukemia/lymphoma 1A, TCL1A
32872_at: Transcription factor 4, TCF4