Gene Learning Association

Our association was born from the intersection of our interests in the development of innovative diagnostic methods and in cancer treatments. Three of our members are recognized researchers who are developing a promising method of diagnosing cancer and other pathologies; other members have embraced this cause by the need to learn about different approaches to treating cancer and to share their knowledge with those who need it. Innovative research We developed a novel approach to gene expression analysis and training of diagnostic and prognostic human pathology models. First, it has to be noted that most of human pathologies affect expression of not just one but rather of many genes. Second, it is then easy to realize that disease might have an impact on many parts of entire gene network. Thus, instead of searching for the specific genes we used gene expression stoichiometries to develop diagnostic and prognostic models for human pathologies. In principle, our holistic approach results in the following key advantages: 1) Significant reduction in stoichiometry signatures models’ dimensionality. 2) Reduction in noise for stoichiometry signatures. As a result, reduced noise in feature space simplifies models training and improves their cross-validation accuracy. 3) Models trained on big data generated by high-throughput RNA-sequencing or microarrays technologies potentially can be transferred to low cost platforms, such as RT-PCR. 4) Last, but not least, our approach may uncover novel therapeutic targets based on analysis of stoichiometry changes in constituents of gene network. Information ressource Considering demographics and aging of human populations cancer is expected to remain as one of the leading mortality-causing disease. Thus, our community service will be to collect information and create an information ressource on innovative anti-cancer cures for those who need it.