從機率分布到貝氏推論，從多重檢定到存活分析——17 個主題帶你打通生物資訊統計的任督二脈，每頁皆附互動模擬、R / Python 雙語程式碼與三題自我檢測。
From probability distributions to Bayesian inference, from multiple testing to survival analysis — 17 modules tying classical statistics to bioinformatics practice, each with interactive simulations, R/Python code, and self-check quizzes.
隨機變數、離散/連續分布、CLT 與 LLN——為 RNA-seq 計數 (NB) 與檢定統計量打下基礎。
Random variables, discrete/continuous distributions, CLT and LLN — the foundation for NB count models and test statistics.
MLE、動差法、Fisher information 與 Cramér-Rao 下界，理解 DESeq2 為何要做 dispersion shrinkage。
MLE, method of moments, Fisher information, Cramér-Rao — why DESeq2 shrinks dispersion.
CI 的正確解讀，BCa bootstrap 補救解析解失效的場景，以及與貝氏 credible interval 的差異。
Correct CI interpretation, BCa bootstrap when analytic SEs fail, contrast with credible intervals.
型一/型二錯誤、檢定力、p-value 誤用——告別「p>0.05 等於沒效果」。
Type I/II errors, power, p-value misuse — escape "p>0.05 means no effect".
t / Welch / 配對 / Wilcoxon / Mann-Whitney / χ² / Fisher / ANOVA / Kruskal——附決策樹。
t / Welch / paired / Wilcoxon / Mann-Whitney / χ² / Fisher / ANOVA / Kruskal — with a decision tree.
OLS 假設、殘差圖、VIF、Cook's distance、HC3 穩健標準誤。
OLS assumptions, residual plots, VIF, Cook's distance, HC3 robust SEs.
lme4、批次效應、部分匯整 (partial pooling)、emmeans 對比。
lme4, batch effects, partial pooling, emmeans contrasts.
logit 連結、OR 解讀、Firth 修正、ROC vs 校準。
logit link, OR interpretation, Firth correction, ROC vs calibration.
連結函數、deviance、過度離散、零膨脹——直通 DESeq2 / edgeR。
Links, deviance, overdispersion, zero-inflation — the bridge to DESeq2 / edgeR.
Beta-Binomial、Gamma-Poisson 等共軛先驗，與後驗預測分布。
Beta-Binomial, Gamma-Poisson conjugate priors, and posterior predictive distributions.
Stan / brms / PyMC、NUTS、R-hat、ESS、divergent transitions。
Stan / brms / PyMC, NUTS, R-hat, ESS, divergent transitions.
HDI / ETI、posterior predictive check、LOO-CV、WAIC。
HDI vs ETI, posterior predictive checks, LOO-CV, WAIC.
James-Stein、limma eBayes、ashr——20,000 基因如何借力借強度。
James-Stein, limma eBayes, ashr — borrowing strength across 20,000 genes.
Bonferroni / BH / Storey q-value / IHW / local FDR；p-value 直方圖診斷。
Bonferroni / BH / Storey q-value / IHW / local FDR; p-value histogram diagnostics.
voom、NB GLM、QL F-test、lfcShrink (apeglm/ashr)。
voom, NB GLM, QL F-test, lfcShrink (apeglm/ashr).
樣本排列 vs 基因排列、CAMERA、fgsea、competitive vs self-contained。
Sample vs gene permutation, CAMERA, fgsea, competitive vs self-contained tests.
Kaplan-Meier、log-rank、Cox PH、Schoenfeld 殘差、競爭風險。
Kaplan-Meier, log-rank, Cox PH, Schoenfeld residuals, competing risks.
17 主題的權威書籍（Wasserman、Casella & Berger、Gelman BDA3）、原始論文、套件文件，附 DOI 連結。
Canonical books (Wasserman, Casella & Berger, Gelman BDA3), original papers, and package docs for all 17 topics — with DOI links.
涵蓋 17 主題的單選、多選、是非題，內建錯題複習、隨機練習、考試模式與多使用者進度。
Single-choice, multi-select, and T/F questions across all 17 topics — with wrong-answer review, exam mode, and multi-user progress tracking.