Résumé
<p><b>OBJECTIVE</b>To formulate recommendations in the evaluation of results of genetic analyses in paternity testing under considering mutations.</p><p><b>METHODS</b>A total of 15 short tandem repeat(STR) loci were employed for this study, which were included CSF1PO, FGA, TH01, TPOX, VWA, D3S1358, D5S818, D7S820, D8S1179, D13S317, D16S539, D18S51, D21S11, PentaD and PentaE. Both 100 cases of true trio and 100 cases of false trio were investigated.</p><p><b>RESULTS</b>The numbers of mismatch alleles in different STR loci were observed in 100 cases of false trio. The different distributions of paternity index were obtained, including the changes of paternity index in each case of true trio under simulated mutations.</p><p><b>CONCLUSION</b>In order to avoid the effect of mutations, the exclusion of paternity was never considered on the basis of a single locus. The threshold values of the combined probability of exclusion and the paternity index were important for both exclusion and inclusion of paternity. The scientific evidence for paternity testing can be obtained when both the combined probability of exclusion and the paternity index meet the threshold values. However, when either the combined probability of exclusion or the paternity index can not meet the threshold values, more genetic markers should be added.</p>