CLASSIF1 Classification

Pretherapeutic Identification of High Risk DLBCL Patients

fig.1: Predictive Data Pattern      > fig.2

- The most discriminatory data pattern (disease classification masks) for the predictive identification of ultimate therapy survivors (S) and nonsurvivors (NS) was selected by the CLASSIF1 data sieving algorithm from the DLBCL. database parameters (Lymphochip, 7399 gene expression array).
- The disease classification mask contains 24 (0.32%) informative parameters for the individualized outcome prediction concerning standard therapy.
- Fourteen gene products correspond to known molecular entities while the 8 gene products concern molecules without known function.
- The above predictive data pattern is significantly different from the communicated prognostic data pattern consisting of 17 parameters with HLA-DPalpha |Hs.914| as the only coincident parameter.
- The predictive pattern comprises 13 increased (+) and 11 decreased (-) parameters for non survivors.

© 2024 G.Valet
1965-2006: Max-Planck-Institut für Biochemie, Am Klopferspitz 18a, D-82152 Martinsried, Germany
Last Update: Apr.3 ,2003