i am trying to solve a structure of a protein (catalytic in nature). i ve got some crystal in  sodium formate. crystal have thin plate morphology and whenever i am trying to diffract them, at certain images spot streaking is observed. so i am trying to merge some data sets as one data set  alone is not sufficient for completeness. this results in a high Rmerge say 0.25 along with 3.2 redundancy and completeness of 98.7 (total) and 89.4 (last bin) with i/sigma >2 in last bin (2.59A). how can i improve my data.  datas are collected in in Cu kalpha anode. ive tried peg 400, glycerol, ehylene glycol, sucrose and sorbitol as cryo supplements with formate but there is no improvement. further the datasets have high mosaicity >1 and cell parameters vary sometime. the space goup is C121. i am using HKL2000 for scaling and integration. i ve even tried seeding but morphology remains the same. the protein is highly pure >95% and homogenous (mass spectro anlysis Maldi). Any suggestions will be helpful.

thanx in advance

rakesh