I am trying to make a phylogenetic tree using GTR model and maximum likelihood method by MEGA. I am using nucleotide sequences of a specific family of proteins from wheat and since its genome sequencing is still under progress there is some missing information (represented by NNNN) and gaps and as a result I get an error every time I use MEGA for tree construction. The error says no common sites found. But I get this error only when I use partial deletion in gap treatment. I do not receive any error if I choose "use all sites" for gap treatment or decrease the site coverage cutoff to ~60% for partial deletion. Can anyone explain to me the difference between the two, I do not want to loose any information. Also I do not want go add any false information too and i am pretty sure the sequences I am using are from one single family of proteins only. Thank You