After the DESEQ2 analysis, the baseMean values from the results range from (0.2 to 87622.00). For the selected results I have used 0.05 FDR cutt off. But, still some baseMeans are as low as 0.2.

Can I use the baseMean cut off as 5.

My concern is if the baseMean is this low , then the log2FC will be high, but may not be real.

Please advice.

The low end of basemean doesn’t mean there are any small pvalues with that basemean. Also, a gene can have a low basemean and be clearly DE. But of course it may make sense to just set a threshold yourself and set independentFiltering=FALSE. You can filter right after estimateSizeFactors() and before running DESeq().