Abstract
The modified logistic regression (Morais-rodrigues et al., 2020) was used to classify breast cancer subtypes using all microarray database samples. A stabilizing term, that allows the assignment of values to αi∗ parameters, was inserted in this methodology and with that, during the derivation there is the insertion of the identity matrix (positive defined) which added to the other semi-defined part, results in a positive defined matrix, which has auto values > 0 and determinant > 0, square matrix is full rank if it is reversible (determinant > 0) , which results in a single solution. In the results it was observed that some genes were located topologically in the extremities after plotting the parameters αi∗, these parameters are related to the expression of genes with a suppressor or oncogenic profile in breast cancer, and with genes not studied yet. Some of these genes were found in gene regulatory networks from the search of Iglesias-Martinez et al. (2016), and S-score values were associated with these genes, negative value S-score is indicative of tumor suppressing or reduced gene activity and the positive value S-score is indicative of oncogene or increased gene activity (de Souza et al., 2014). In view of the importance of these genes, this article provides a literary review of them.