The background error rate of pyrosequencing was calculated with a plasmid containing a cloned HCV sequence (pCV-J4L6S)[24] and the read number for the plasmid is also shown in Table 2. Though seven runs of the plasmid produced 2277–7000 reads, with an average of 5448 reads, there was no background buy Gefitinib error at amino acid (a.a.) 31, 32 or 93 in NS5A. Because the background error rate was 0% at each position, the presence of mutations at 0.1% or higher was considered
to be significant, based on the 95% confidence interval (0–0.1%) calculated for 0% in 2227 reads. The background error rate coincided almost exactly with the background error rate obtained in our recent study.[23] The baseline characteristics of the 110 patients are shown in Table 1. The data for viral factors (core a.a. 70, core a.a. 91, NS5A-ISDR and NS5A-IRRDR) in the table were obtained by direct sequencing as described in Methods. As shown in the table, there were significant
differences among the three groups in aspartate aminotransferase, alanine aminotransferase, γ-glutamyltransferase, α-fetoprotein, core a.a. 70 and IL28B SNP (rs8099917). Meanwhile, there was no significant difference in background factors of age and sex or factors associated with liver fibrosis such as platelets and albumin. Because previous reports showed that L31M/V/F, P32L and Y93H are resistance mutations in NS5A of genotype 1b HCV, the presence of these mutations Pictilisib solubility dmso was analyzed by deep sequencing. Table 3 shows the rate of NS5A resistance mutations at a.a. 31, 32 and 93. At a.a. 32, no mutation was found in any of the 110 patients. Regarding a.a. 31, resistance mutations (L31M/V/F) were observed in 13 of the 110 patients (11.8%) and, despite no significant difference, tended to occur more frequently in the relapser group and naïve group than in the null group. Meanwhile, the a.a. 93 resistance mutation (Y93H) was observed in 34 of the 110 patients Selleckchem Rucaparib (30.9%) and, despite no significant
difference, also tended to occur more frequently in the naïve and relapser groups than in the null group. Simultaneous a.a. 93 and 31 resistance mutations were observed in only four of 110 patients (3.6%) and these four patients all belonged to the naïve group. More detailed deep sequence results for the four patients with simultaneous mutation of L31M/V/F and Y93H are shown in Table 2. Although the substitution rate of L31M/V/F in these patients was low, all isolates with L31M/V/F also featured the Y93H change. Figure 1 show the mutation rates of L31M/V/F and Y93H in each patient. One bar indicates the resistance mutation rate in one patient, obtained by deep sequencing. It was found that minor viral populations that were not detected by direct sequencing could be detected by deep sequencing.