Hi,
(1) I just wonder if this tool is just the step to select features before building a model.
(2) In addition to clinical end points, if there are difference in other factors between two groups of samples, such as family history, smoking or not, female or male, can this tool be used to select features to build a model to predict those meta information from another test set where we didn't collect those meta information?
(3) Or can it be used to select biomarker genes in a cell type in a single-cell dataset, including positive and negative markers?
Thank you.
It is mainly focus on that yes. We are mainly using Stabl to select features.
2-3. It depends on the nature of your data. Consider it as an estimator in a multivariate analysis. In your example, a confounder analysis might be interesting. In general, if your data are in a wide format, it can be used to select features.
Hi, (1) I just wonder if this tool is just the step to select features before building a model. (2) In addition to clinical end points, if there are difference in other factors between two groups of samples, such as family history, smoking or not, female or male, can this tool be used to select features to build a model to predict those meta information from another test set where we didn't collect those meta information? (3) Or can it be used to select biomarker genes in a cell type in a single-cell dataset, including positive and negative markers? Thank you.