it would be great if one could use normalized and/or scaled data as input to InferCNV. Over the last years, a lot of batch-effect/integration/harmonization or even deconvolutional tools have been developed, that help dealing with more complex datasets. Typically, they return scaled data. Some sort of manual control over the input could thereby lift the potential of InferCNV. I know that this topic has already been here a long time ago, but this is becoming more and more important and could be an interesting future direction.
Dear Dev-Team,
it would be great if one could use normalized and/or scaled data as input to InferCNV. Over the last years, a lot of batch-effect/integration/harmonization or even deconvolutional tools have been developed, that help dealing with more complex datasets. Typically, they return scaled data. Some sort of manual control over the input could thereby lift the potential of InferCNV. I know that this topic has already been here a long time ago, but this is becoming more and more important and could be an interesting future direction.