Feature selection for multi-class classification problem.

Question

[Shakunthala Natarajan]

I have to classify my samples into 6 different classes using gene expression data as features. I plan to use a logistic regression based ML model or XGboost for the classification. I would like to adopt a feature selection method that captures the highly variable genes. I want to do this in python. I want to find an equivalent approach in python for the BrenneckeGetVariableGenes function of M3Drop R package. Are there any suitable modules or approaches in python that you would recommend for this use case? Thank you!