An interlaboratory comparison of typing results for Short Tandem Repea
ts (STRs) at the GenBank loci HUMCSF1PO, HUMTPOX, HUMTH01, and HUMVWFA
31 using the ''CTT triplex'' and ''CTTv quadruplex'' has been evaluate
d. These STRs all have a nominal four basepair (bp) repeat. Seven diff
erent samples were distributed to 41 laboratories. The 34 laboratories
that returned results used a wide variety of analytical systems. Comp
arable results were obtained for all samples at all loci when results
were reported as an allelic name. Raw sizing results obtained from int
ernal-lane sizing standards differed by nearly five bp at some loci. M
any different factors contribute to this observed sizing variability,
including choice of sizing standards and matrix composition. Although
sizing results can be made more comparable by locus-specific offsets o
r calibration to a comprehensive set of alleles at each locus, samples
typed to the allelic name can now be validly compared regardless of a
nalytical method. Interlaboratory comparison of raw allelic size remai
ns problematic.