i just need to know is it necessary to validate bam after marking PCR duplicate using MarkDuplicates for downstream analysis in GATK
It's not necessary but it's a good idea to do it, so that if anything's wrong with your data, you will find out before you invest a lot of time and effort analyzing it...
thanks Geraldine_VdAuwera actually when i validate i get first NM tag error
Record 86, Read name 195_1700_2009, NM tag (nucleotide differences) in file  does not match reality 
I AM STUCK on validation step
Errors like NM tags and mate pair information are usually not important, so you can tell ValidateSAMFile to skip those. Please see the Picard Tools documentation for more details and for help interpreting the results. These are not our tools so we can't provide detailed support for them. I think there is a mailing list for Picard tools support. Good luck!
Thank you so much Geraldine_VdAuwera you are a very co-operative person. let me try
if PCR percent duplication in our sample is 0.171634 can we ignore in GATK pipeline for variant calling
It depends what you mean by "ignore in GATK pipeline for variant calling". If you mark duplicates according to our Best Practices recommendations, the duplicates will be ignored (which is the right thing to do).
i mean is this percentage is large enough to mark duplicates
Yes you should always mark duplicates, even if the proportion is not very high. In variant analysis, every little bit of cleanup helps to separate the signal from the noise.
ok thank you so much