scRNA-based variants ?

[ahy1221]

It seems that would be a powerful tool for the scRNA-seq data.
One of main concern is that acquiring reliable variants set for scRNA seq is challenge due to the high false positive and the limited sensitivity. I wonder that do you have any reliable/recommended pipeline to get variants from scRNA-seq data? It would good to have a vignette about preparing variant data I guess

[davismcc]

Hi ahy1221,

Thanks for your interest in cardelino! You are right that obtaining a set of reliable variants for clonal inference for scRNA-seq data is a challenge (not such a problem for donor ID, where we have good reference sets of germline variants).

In our applications so far, we have defined somatic variant sites using other data (specifically, whole-bulk whole-exome data) and extracted the reference and alternative allele counts at these positions as input to the cardelino clone ID function. Defining somatic variant sites from scRNA-seq dat alone would be very appealing, but this is very far from being a solved problem.

I agree about providing a vignette about preparing variant data for use with cardelino. Although the pre-processing falls outside the scope of the package itself, it would be valuable to show people some strategies for how they might do it. We'll aim to add such a vignette to the package in the near future.

Best,
Davis