Vivianstats / scImpute

Accurate and robust imputation of scRNA-seq data
https://www.nature.com/articles/s41467-018-03405-7
91 stars 34 forks source link

Parameter Selection #12

Closed radio1988 closed 6 years ago

radio1988 commented 6 years ago

With scImpute, how can we select parameters for drop_thre and Kclusters? Should we always use the default? Thanks!

Vivianstats commented 6 years ago

Hello,

For drop_thre, I suggest that you start with 0.5 and in most cases, it works pretty well.

As for Kclusters, you can perform an initial clustering of the raw data, and hopefully, that gives you a vague idea of the approximate number of clusters in your data. A smaller Kclusters makes scImpute more conservative.

Best, Vivian

radio1988 commented 6 years ago

Thanks for the explanation!