Results were quite positive and the presence of a 20% molar contaminant did not interfere with manual sequence assignment accuracy. An average of 30 residues were correctly assigned for the primary sequence (a 40 residue peptide) and not a single positive assignment was wrong due to the presence of the contaminant. The overall assignment accuracy of positive calls was 95%, and 20 of a total 48 facilities made no errors in positive sequence assignments. Only in a few cases were a substantial number of errors made, with the most common problem being slipping out of phase.
Computerized data assignments were much less precise than manual assignments. The data suggested that accuracy levels above 90% could be obtained if only those automatic calls were considered that had pmol ratios above 20.
Single aliquots of ABRF-89SEQ were also sent by the Executive Committee to each sequence instrument manufacturer as an unknown. Porton Instruments was the only manufacturer that returned the data prior to release of the sequence. They assigned all of the first 39 residues correctly for the primary component. In addition, 38 positive assignments were made in the secondary sequence and all were correct. Also, one of eight tentative assignments was correct. As these numbers imply, tentative assignments were made for both peptides beyond where the peptides actually ended. This problem of "over assignment" was a}so seen in data from a substantial number of core facilities.
[ABRF does not endorse any type of instrument.]
Return to the The ABRF Home Page
Created: 10th August 1995
Last modified: 10th August 1995