The test sample distributed this year, ABRF-93SEQ, was designed by rearranging the sequence of the first ABRF test peptide, STD-I, released in 1988. ABRF93SEQ contained no unusual residues, periodic alanines for evaluating repetitive yield, and several potentially problematic sites, notably Gl, W2, C5, W7, R19-T20-R21, and H38:
GWVACMWSSA RDFHPAYIRT RAYSMVPARL EEVAGIIEIAK
All 274 ABRF member laboratories received 50 pmol of this sample, and the first 80 responses were analyzed by the Committee for this year's report.
Altogether, 87% of the reported sequence assignments were correct, and 91% of the positive assignments were correct. An average of 28 assignments were made for this 40 residue peptide, with 25 correct and 4 tentative calls. C5, W7, W2, and G1 were the most troublesome residues, and to a lesser extent, all S, H, and R residues in the test sample. Performance with C and W separated most responses into two groups, one that identified all three residues consistently and another that did not identify them at all. Difficulty with R assignments was emphasized at the segment R 19-T20-R21, where T was called more successfully than either R residue.
Three laboratories assigned the entire sequence correctly, and one of these made forty positive correct calls, a rare achievement in the history of these studies. A common feature of the three best responses was the review of sequence data by more than one individual and, in each case, at least one reviewer with more than 10 years experience. There were 34 groups with no assignment errors. On the other hand, 16 groups had five or more positive incorrect assignments, and 4 groups made more incorrect than correct positive calls.
Comparison of results obtained with ABRF-93SEQ (50 pmol) and STD-I (100 pmol) showed overall sequencing performance has not improved in the past five years. This may be due to differences in the level of experience of the laboratories participating in the two studies or to the lower amount of sample provided in the 1993 study. It may also indicate that improvements in instrumentation and methodology have not been readily and effectively implemented in many laboratories. In addition, the five previous studies conducted by this Committee have shown that different instrument models from the same vendor perform similarly in the field, suggesting that instrumentation improvements have not significantly impacted sequencing performance.
This Committee will attempt to provide the ABRF News with a listing of improvements that all protein sequencing groups should consider evaluating in their own laboratories. For any group dissatisfied with their performance in this year's study, we will also provide a previously unreleased test sample.
The 1994 Committee members are Philip Andrews, Jay Gambee, Gregory Grant, Sheenah Mische, and John Rush, all returning from the 1993 Committee. They will be joined by Karen De Jongh and Barbara Merrill, who replace the departing senior members Dan Crimmins and David Speicher.
Return to the The ABRF Home Page