Analysis Methods
DDCt (Livak method)
The Delta-Delta Ct method (Livak & Schmittgen, 2001) calculates relative fold-change in gene expression between experimental and control samples, normalized to one or more reference genes.
When multiple reference genes are selected, AnnealIQ uses the geometric mean for normalization — the recommended approach for multi-reference gene studies.
Error propagation uses SD and SEM across biological replicates. The 2^(–ΔΔCt) formula assumes near-100% and equal amplification efficiency for target and reference genes. If your primer efficiencies differ substantially from 100%, consider the Pfaffl method.
Pfaffl method (efficiency-corrected)
The Pfaffl method corrects for differences in amplification efficiency between target and reference genes. Enter efficiencies manually or calculate them from a standard curve.
Error bars are computed in the log domain and back-transformed, producing correctly asymmetric bars on the linear scale. An info tooltip explains this methodology.
AnnealIQ recommends Pfaffl automatically when primer efficiencies fall outside 90–110%.
AI method selection
The AI evaluates your data and context to recommend DDCt or Pfaffl. A method selection card appears in the chat showing the chosen method, the reasoning, and an alternative you can switch to.
If you’ve entered primer efficiency values outside 90–110%, the AI will recommend Pfaffl and explain why. You can always override the recommendation.
95% bootstrap confidence intervals
AnnealIQ computes nonparametric bootstrap confidence intervals (1,000 iterations) of biological replicates in log2 space (Efron & Tibshirani, 1993). CIs are available for both DDCt and Pfaffl methods.
In the results table, a [lower, upper] column shows the CI range. A blue alert icon appears when the CI crosses 1.0 (indicating no statistically significant change). CIs require at least 2 biological replicates.
When you select "95% CI" as the error bar type on the chart, the actual bootstrap CIs are used (asymmetric on linear scale). The chart legend shows "95% Bootstrap CI (1000 iterations, n=N)".
Parameter confirmation
Before running any analysis, AnnealIQ shows a parameter confirmation card listing: reference gene(s), control group, analysis method, and any excluded replicates.
Each parameter has a confidence badge: mentioned (you stated it explicitly), inferred (derived from your data or description), or assumed (default choice). Review the inferred and assumed parameters before confirming.