Machine learning classifier of tumor cells
ikarus is a stepwise machine learning pipeline that tries to cope with a task of distinguishing tumor cells from normal cells. Leveraging multiple annotated single cell datasets it can be used to define a gene set specific to tumor cells. First, the latter gene set is used to rank cells and then to train a logistic classifier for the robust classification of tumor and normal cells. Finally, sensitivity is increased by propagating the cell labels based on a custom cell-cell network. ikarus is tested on multiple single cell datasets to ascertain that it achieves high sensitivity and specificity in multiple experimental contexts.
| System | Target | Derivation | Build status |
|---|---|---|---|
| x86_64-linux | /gnu/store/r7bj0dn4j1iws8kda99n6g22jwgmqbic-python-ikarus-0.0.2.drv |
| Linter | Message | Location |
|---|---|---|
| No lint warnings ✓ | ||