The data generated from Smart-seq2 single-cell RNA-Seq protocol seems to have very unbalanced gene expression values. For example, if you did the differential gene expression test, you would find out many genes are only present in one sample group. Is this normal?
I wonder if there are some specific pipelines including some normalization procedure to deal with these data?
Or alternatively, can we use the pipeline of handling the bulk RNA-Seq such as Tophat2 + Cufflink + Cuffdiff or STAR + Featurecount + DESeq2 to analyze the data from Smart-seq2 single-cell RNA-Seq protocol?