Missing values in expression data matrix

[Amative]

Hello,

I have two expression data matrices (Reads Count and RPKM) both for the same set of samples. I have a lot of missing values in both. I am wondering if anyone can guide me to the correct way to deal with the missing expression values. Some if the genes missing expression in all samples, while some missing ( 1, 2, 3 ... etc). Should I get rid of those genes with more missing values than a specific threshold or what?

Thanks

[saulbishop]

Sorry, the data matrix you just mentions is not the barcode type "Data Matrix", isn't it?