GEO Signature Extractor
- 0 Collaborators
This project aims at developing a gene expression signature extractor on GEO studies. It can be used to construct the signatures for the subject (e.g. gene, drug, disease) in each study. The signatures can be further used to build the similarity network for inferring the associations among those subjects. GESgnExt consists of multiple components for signature construction. It provides a set of functions to evaluate the performance of different methods for each component. The most recommended model for the classification components is UDT-RF and the one for the clustering component is Kallima. The trained models can be adopted for predictions of subject categorization and control/perturbation sample discrimination. It also provides several utility functions to manipulate the dataset and post-process the results. ...learn more
Project status: Under Development