Background: Precise determination of amplification efficiency is critical for reliable conversion of within-sample changes in fluorescence occurring on a logarithmic scale to between-sample differences in DNA content occurring on a linear scale. This endeavor is especially challenging for the telomere length (TL) quantitative-PCR (qPCR) assay, where amplification efficiency can vary between reactions targeting telomeric repeats (T) and those targeting a single-copy gene (S) to calculate TL as the T/S ratio.

Methods: We compared seven different approaches toward estimating amplification efficiency, including the standard-curve method utilized by the qPCR instrument software, and alternative approaches which estimate efficiency on a reaction-by-reaction basis using the stand-alone program LinRegPCR. After calculating T/S ratios using efficiency estimates from each approach (N = 363), we tested their relative performance on metrics of assay precision and correlates of external validity including chronological age (age range = 1–72 years), across tissues within-person (leukocyte-buccal), and between parents and offspring.

Results: Estimated amplification efficiency for telomere reactions was significantly lower than estimates for single-copy gene reactions. Efficiency estimates for both reaction sets were significantly higher when estimated with the standard-curve method utilized by the qPCR instrument relative to estimates reconstructed during the log-linear phase with LinRegPCR. While estimates of single-copy gene efficiency reconstructed using LinRegPCR measured within 90% of perfect exponential doubling (E = 1.92), estimates generated using the standard-curve method were inflated beyond 100% (E = 2.10–2.12), indicating poor fidelity. Despite differences in raw value, TL measurements calculated with LinRegPCR efficiency estimates exhibited similar relationships with external validity correlates to measurements generated using the qPCR instrument software.

Conclusion: Since methods to estimate amplification efficiency can vary across qPCR instruments, we suggest that future analyses empirically consider external methods of efficiency calculations such as LinRegPCR, and that already generated data be re-analyzed to glean possible improvements.

背景：精确测定扩增效率对于将样品内以对数尺度发生的荧光变化可靠地转换为以线性尺度发生的样品间 DNA 含量差异至关重要。这项工作对于端粒长度 (TL) 定量 PCR (qPCR) 测定尤其具有挑战性，其中针对端粒重复 (T) 的反应和针对单拷贝基因 (S) 的反应之间的扩增效率可能会有所不同，以将 TL 计算为 T /S 比率。

方法：我们比较了七种不同的估计扩增效率的方法，包括 qPCR 仪器软件使用的标准曲线方法，以及使用独立程序 LinRegPCR 在逐个反应的基础上估计效率的替代方法。使用每种方法的效率估计计算 T/S 比率后（氮= 363），我们测试了它们在测定精度指标和外部有效性相关性方面的相对表现，包括实际年龄（年龄范围 = 1-72 岁）、人体内组织（白细胞-口腔）以及父母和后代之间的关系。

结果：端粒反应的估计扩增效率显着低于单拷贝基因反应的估计扩增效率。相对于使用 LinRegPCR 在对数线性阶段重建的估计值，使用 qPCR 仪器使用的标准曲线方法估计时，两个反应组的效率估计值显着更高。而使用 LinRegPCR 重建的单拷贝基因效率估计值在完美指数倍增的 90% 以内（乙= 1.92），使用标准曲线方法生成的估计值夸大超过 100%（乙= 2.10–2.12)，表明保真度较差。尽管原始值存在差异，但使用 LinRegPCR 效率估计计算的 TL 测量值与使用 qPCR 仪器软件生成的测量值相关的外部有效性表现出类似的关系。

结论：由于估计扩增效率的方法可能因 qPCR 仪器而异，因此我们建议未来的分析根据经验考虑外部效率计算方法（例如 LinRegPCR），并重新分析已生成的数据以收集可能的改进。