From 4e6398535d2811fc7ca9faab96e45a9fdbd12668 Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Mon, 20 Apr 2026 22:54:19 +0200
Subject: [PATCH 01/13] Basic schema

---
 .../faqs/index.md                             | 138 +++++++
 .../tutorial.bib                              | 377 ++++++++++++++++++
 .../tutorial.md                               | 358 +++++++++++++++++
 3 files changed, 873 insertions(+)
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/index.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.bib
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/index.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/index.md
new file mode 100644
index 00000000000000..821d165ce527a4
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/index.md
@@ -0,0 +1,138 @@
+# Frequently Asked Questions about TSENAT and Transcriptome Complexity Analysis
+
+## Conceptual Questions
+
+### What's the difference between Tsallis entropy and Shannon entropy?
+
+**Shannon entropy** (q=1 in Tsallis framework) treats all isoforms equally. **Tsallis entropy** adds a parameter `q` that acts as a sensitivity dial:
+- When q < 1: emphasizes **rare isoforms**
+- When q = 1: recovers Shannon entropy (balanced)
+- When q > 1: emphasizes **abundant isoforms**
+
+This allows you to examine complexity at different biological scales in a single analysis.
+
+### Why is transcriptome complexity important if DESeq2 already detects changes?
+
+DESeq2 and edgeR measure **total gene abundance** changes. They miss situations where:
+- A gene maintains constant total expression but **reorganizes its isoforms**
+- Rare isoforms become more common (or vice versa) without affecting overall counts
+- The **diversity** of the isoform landscape changes
+
+These are valid biological signals (isoform switching via splicing regulation) that complement abundance-based methods.
+
+## Technical Questions
+
+### What sample size do I need?
+
+TSENAT assumes **≥6-8 samples per group** for paired/longitudinal designs:
+- Smaller samples are underpowered to detect entropy shifts
+- Larger samples increase power to detect subtle complexity changes
+- Biological replicates are essential (not technical replicates)
+
+### How do I choose the q-spectrum (range and granularity)?
+
+**Standard configuration**: q = 0 to 2.0 with Δq = 0.05 (41 q-values)
+- Range: 0-2 covers extreme rare-emphasis to abundant-emphasis
+- Granularity: Δq=0.05 provides good resolution without computational burden
+- For visualization: Can downsample to q = (0, 0.5, 1, 1.5, 2) for quick exploration
+
+### Does read depth affect entropy estimates?
+
+Yes. TSENAT uses **automatic pseudocount selection** to handle this:
+- Pseudocounts scale with library size
+- Ensures entropy estimates are comparable across samples with different sequencing depths
+- Breaks ties in rank-based validation methods
+- Standard practice in RNA-seq analysis
+
+### What's the difference between "consistent" vs. "scale-dependent" transcript switching?
+
+- **Consistent switching**: Same transcripts show high delta-influence values across all q-values
+  - Robust, scale-independent splicing shift
+  - Suggests coordinated regulatory mechanism
+  
+- **Scale-dependent switching**: Different transcripts matter at different q-values
+  - Regulatory complexity varies by scale
+  - Different mechanisms for rare vs. abundant isoforms
+
+## Methodological Questions
+
+### What does "q × condition interaction" mean?
+
+It means the effect of **condition on entropy depends on which q-value** you examine:
+- At q=0.5, one condition might have higher entropy
+- At q=2, the pattern might reverse
+- This reveals **scale-dependent** biological signals
+
+### What's the advantage of Tsallis divergence over fold-change?
+
+**Fold-change**:
+- Simple 1:1 ratio between conditions
+- Misses complexity patterns
+- Unbounded (can be arbitrarily large)
+
+**Tsallis divergence**:
+- Information-theoretic distance between distributions
+- Captures pattern type (rare-driven, balanced, abundant-driven)
+- Symmetric [0] to asymmetric, bounded interpretability
+- Directly quantifies isoform complexity differences
+
+## Practical Workflow Questions
+
+### Where do I get transcript-level counts for this analysis?
+
+TSENAT expects **SALMON or Kallisto output** including:
+- NumReads: raw fragment counts per transcript
+- TPM: length- and library-normalized estimates
+- EffectiveLength: read-length corrected transcript length
+
+These are standard outputs from pseudoalignment-based quantification.
+
+
+## Interpretation Questions
+
+### A gene shows significant q×condition interaction but small effect size—should I report it?
+
+Guidelines for meaningful results:
+- **Effect size (Tsallis divergence) D > 0.1**: ~10% information divergence (notable)
+- **D > 0.2**: Substantial divergence (recommended threshold)
+- **Small D with p < 0.05**: May reflect statistical noise; prioritize high-effect genes
+
+Report both for completeness:
+- High-effect genes (D > 0.2): Strong biological signal
+- Borderline genes (D ~ 0.1, p < 0.05): Mention as candidates for validation
+
+
+## Troubleshooting
+
+### My results show no significant genes. What could be wrong?
+
+1. **Sample size too small**: Need ≥6-8 per group (check design)
+2. **Entropy differences too subtle**: Try visualizing q-curves manually
+3. **Filtering too stringent**: Relaxing filters may reveal additional patterns
+4. **Method choice**: Try alternative statistical method (GAM vs. LMM vs. SRH)
+5. **Data quality**: Run M-estimation QC; check for outlier samples
+
+### Some q-values give NA or NaN results. Why?
+
+Possible causes:
+1. **Zero counts**: Pseudocount selection should handle; check filtering
+2. **Numerical instability**: Very small divergence values can underflow
+3. **Method-specific**: SRH (rank test) may have issues with many tied values
+
+Solution: Ensure pseudocount regularization is enabled; try alternative statistical method.
+
+### Q-curves show erratic patterns. Is this normal?
+
+**Expected**: Small fluctuations in entropy across q-values
+**Concerning**: Large spikes or reversals
+- May indicate normalization issues
+- Check: Are housekeeping genes (GAPDH, ACTB) showing balanced patterns?
+- If not: Revisit pseudocount and filtering parameters
+
+### Can I use TSENAT with very long reads (PacBio, Nanopore)?
+
+Conceptually yes, but:
+- TSENAT expects pseudoalignment-based quantification (SALMON, Kallisto)
+- Long reads typically use alignment-based tools (Minimap2, etc.)
+- Workflow would need adaptation to parse alignment-based counts
+- Overall approach (entropy framework) remains valid
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.bib b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.bib
new file mode 100644
index 00000000000000..e24318a8c641c7
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.bib
@@ -0,0 +1,377 @@
+@article{adami2004information,
+  title={Information theory in molecular biology},
+  author={Adami, Christoph},
+  journal={Physics of Life Reviews},
+  volume={1},
+  number={1},
+  pages={3--22},
+  year={2004},
+  publisher={Elsevier},
+  url={https://doi.org/10.1016/j.plrev.2004.01.002}
+}
+
+@article{alomani2023properties,
+  title={Further Properties of Tsallis Entropy and Its Application},
+  author={Alomani, Ghazi and Kayid, Mohamed},
+  journal={Entropy},
+  volume={25},
+  number={2},
+  pages={199},
+  year={2023},
+  publisher={MDPI},
+  url={https://doi.org/10.3390/e25020199}
+}
+
+@article{anastasiadis2012entropy,
+  title={Entropy Properties and Multiple Tsallis Distributions},
+  author={Anastasiadis, A},
+  journal={Entropy},
+  volume={14},
+  pages={174--176},
+  year={2012},
+  publisher={MDPI},
+  url={https://doi.org/10.3390/e14020174}
+}
+
+@article{bajic2024information,
+  title={Information Theory, Living Systems, and Communication Engineering},
+  author={Bajic, Dragan},
+  journal={Entropy},
+  volume={26},
+  number={5},
+  pages={430},
+  year={2024},
+  publisher={MDPI},
+  url={https://doi.org/10.3390/e26050430}
+}
+
+@article{bartal2022progress,
+  title={Progress in and Opportunities for Applying Information Theory to Computational Biology and Bioinformatics},
+  author={Bartal, Aryam and Jagodnik, Kathleen M},
+  journal={Entropy},
+  volume={24},
+  number={7},
+  pages={925},
+  year={2022},
+  publisher={MDPI},
+  url={https://doi.org/10.3390/e24070925}
+}
+
+@article{benjamini1995controlling,
+  title={Controlling the false discovery rate: a practical and powerful approach to multiple testing},
+  author={Benjamini, Yoav and Hochberg, Yosef},
+  journal={Journal of the Royal Statistical Society B},
+  volume={57},
+  number={1},
+  pages={289--300},
+  year={1995}
+}
+
+@article{cao2017comprehensive,
+  title={Comprehensive single-cell transcriptional profiling of a multicellular organism},
+  author={Cao, Junyue and Packer, Jonathan S and Ramani, Vijay and Cusanovich, Darren A and Huynh, Chau and Daza, Riza and Qiu, Xiaojie and Lee, Hannah and Nguyen, Scott N and Orzechowski, Rachel and others},
+  journal={Science},
+  volume={357},
+  pages={661--667},
+  year={2017},
+  publisher={American Association for the Advancement of Science},
+  url={https://doi.org/10.1126/science.aam8940}
+}
+
+@book{chakraborty2019introductory,
+  title={Introductory Time Series Analysis},
+  author={Chakraborty, Tapas R},
+  publisher={Indian Statistical Institute, Kolkata},
+  year={2019}
+}
+
+@article{chanda2020information,
+  title={Information Theory in Computational Biology: Where We Stand Today},
+  author={Chanda, Pritam and Costa, Elena and Hu, Jing and Sukumar, Suresh and Van Hemert, James and Walia, Ravi},
+  journal={Entropy},
+  volume={22},
+  number={6},
+  pages={627},
+  year={2020},
+  publisher={MDPI},
+  url={https://doi.org/10.3390/e22060627}
+}
+
+@article{chao2010phylogenetic,
+  title={Phylogenetic Diversity Measures Based on Hill Numbers},
+  author={Chao, Anne and Chiu, Chun-Huo and Jost, Lou},
+  journal={Philosophical Transactions of the Royal Society B},
+  volume={365},
+  number={1558},
+  pages={3599--3609},
+  year={2010},
+  publisher={Royal Society},
+  url={https://doi.org/10.1098/rstb.2010.0272}
+}
+
+@book{cover2006elements,
+  title={Elements of Information Theory},
+  edition={2nd},
+  author={Cover, Thomas M and Thomas, Joy A},
+  publisher={Wiley-Interscience},
+  year={2006}
+}
+
+@book{efron1993introduction,
+  title={An Introduction to the Bootstrap},
+  edition={2nd},
+  author={Efron, Bradley and Tibshirani, Robert J},
+  publisher={Chapman and Hall},
+  year={1993}
+}
+
+@article{erhard2018improved,
+  title={Improved Ribo-seq puromycin target reliability using Bayesian nonparametrics},
+  author={Erhard, Florian and Hense, Benedikt and Jafari, Minoo and Siebourg-Polster, Julia and D{\"o}lken, Lars and Zimmer, Ralf},
+  journal={Bioinformatics},
+  volume={34},
+  number={12},
+  pages={2096--2102},
+  year={2018},
+  publisher={Oxford University Press}
+}
+
+@article{ernst2004permutation,
+  title={Permutation Methods: A Basis for Exact Inference},
+  author={Ernst, Michael D},
+  journal={Statistical Science},
+  volume={19},
+  number={4},
+  pages={676--685},
+  year={2004}
+}
+
+@article{furuichi2006information,
+  title={Information theoretical properties of Tsallis entropies},
+  author={Furuichi, Shigeru},
+  journal={Journal of Mathematical Physics},
+  volume={47},
+  pages={023302},
+  year={2006},
+  publisher={AIP Publishing},
+  url={https://doi.org/10.1063/1.2165744}
+}
+
+@article{gandrillon2021entropy,
+  title={Entropy as a Measure of Variability and Stemness in Single-Cell Transcriptomics},
+  author={Gandrillon, Olivier and Gaillard, Monique and Espinasse, Therry and Garnier, Nicolas B and Dussiau, Cyril and Kosmider, Olivier and Sujobert, Pierre},
+  journal={Entropy},
+  volume={24},
+  number={1},
+  pages={93532},
+  year={2021},
+  publisher={MDPI},
+  url={https://doi.org/10.3390/e24010018}
+}
+
+@article{gao2019uncertainty,
+  title={Uncertainty Measure Based on Tsallis Entropy in Evidence Theory},
+  author={Gao, Xin and Tsai, Sheng-Ben and Liu, Fang and Pan, Lingyun and Deng, Yong},
+  journal={International Journal of Intelligent Systems},
+  volume={34},
+  number={6},
+  pages={1626--1647},
+  year={2019},
+  publisher={Wiley},
+  url={https://doi.org/10.1002/int.22185}
+}
+
+@article{golomb2026information,
+  title={An Information Content Principle Explains Regulatory Patterns},
+  author={Golomb, Reuven and Yoles, Maya and Fishilevich, Sima and Cohen, Orit and Savariego Peled, Elena and Dahary, David and others},
+  journal={bioRxiv},
+  pages={2026.02.19.706555},
+  year={2026},
+  publisher={bioRxiv},
+  url={https://doi.org/10.1101/2026.02.19.706555}
+}
+
+@book{hyndman2018forecasting,
+  title={Forecasting: Principles and Practice},
+  edition={2nd},
+  author={Hyndman, Rob J and Athanasopoulos, George},
+  publisher={OTexts},
+  year={2018},
+  url={https://otexts.com/fpp2/}
+}
+
+@article{jost2006entropy,
+  title={Entropy and Diversity},
+  author={Jost, Lou},
+  journal={Oikos},
+  volume={113},
+  number={2},
+  pages={363--375},
+  year={2006},
+  publisher={Wiley},
+  url={https://doi.org/10.1111/j.2006.0030-1299.14714.x}
+}
+
+@article{jose2013application,
+  title={Application of ARIMA(1,1,0) Model for Predicting Time Delay of Search Engine Crawlers},
+  author={Jose, J and Lal, P S},
+  journal={Informatica Economică},
+  volume={17},
+  number={4},
+  pages={26--39},
+  year={2013},
+  url={https://doi.org/10.12948/issn14531305/17.4.2013.03}
+}
+
+@article{kullback1951information,
+  title={On information and sufficiency},
+  author={Kullback, Solomon and Leibler, Richard A},
+  journal={Annals of Mathematical Statistics},
+  volume={22},
+  number={1},
+  pages={79--86},
+  year={1951},
+  url={https://doi.org/10.1214/aoms/1177729694}
+}
+
+@article{nijman2020perturbation,
+  title={Perturbation-driven entropy as a source of cancer cell heterogeneity},
+  author={Nijman, Sloan Martin B},
+  journal={Trends in Cancer},
+  volume={6},
+  number={6},
+  pages={454--462},
+  year={2020},
+  publisher={Elsevier},
+  url={https://doi.org/10.1016/j.trecan.2020.02.016}
+}
+
+@article{phipson2010permutation,
+  title={Permutation P-values should never be zero},
+  author={Phipson, Belinda and Smyth, Gordon K},
+  journal={Statistical Applications in Genetics and Molecular Biology},
+  volume={9},
+  number={1},
+  pages={Article 39},
+  year={2010}
+}
+
+@article{ramirez2016determining,
+  title={Determining the Entropic Index q of Tsallis Entropy in Images through Redundancy},
+  author={Ram{\'i}rez-Reyes, A and Hern{\'a}ndez-Montoya, A R and Herrera-Corral, G and Dom{\'i}nguez-Jim{\'e}nez, I},
+  journal={Entropy},
+  volume={18},
+  number={8},
+  pages={299},
+  year={2016},
+  publisher={MDPI},
+  url={https://doi.org/10.3390/e18080299}
+}
+
+@article{re2014generalization,
+  title={Generalization of Entropy Based Divergence Measures for Symbolic Sequence Analysis},
+  author={R{\'e}, Mar{\'i}a A and Azad, Rahul K},
+  journal={PLoS ONE},
+  volume={9},
+  number={4},
+  pages={e93532},
+  year={2014},
+  publisher={Public Library of Science},
+  url={https://doi.org/10.1371/journal.pone.0093532}
+}
+
+@article{sason2022divergence,
+  title={Divergence Measures: Mathematical Foundations and Applications in Information-Theoretic and Statistical Problems},
+  author={Sason, Igal},
+  journal={Entropy},
+  volume={24},
+  number={5},
+  pages={712},
+  year={2022},
+  publisher={MDPI},
+  url={https://doi.org/10.3390/e24050712}
+}
+
+@article{seweryn2020application,
+  title={Application of Information Theoretical Approaches to Assess Diversity and Similarity in Single-Cell Transcriptomics},
+  author={Seweryn, Michał T and Pietrzak, Mateusz and Ma, Qing},
+  journal={Computational and Structural Biotechnology Journal},
+  volume={18},
+  pages={1830--1837},
+  year={2020},
+  publisher={Elsevier},
+  url={https://doi.org/10.1016/j.csbj.2020.05.006}
+}
+
+@article{shannon1948mathematical,
+  title={A Mathematical Theory of Communication},
+  author={Shannon, Claude E},
+  journal={The Bell System Technical Journal},
+  volume={27},
+  number={3--4},
+  pages={379--423},
+  year={1948}
+}
+
+@book{shiner2002entropy,
+  title={Entropy and Entropy Generation: Fundamentals and Applications},
+  author={Shiner, Jeffrey S and Emelyanova, Natalya A and Gafarov, Farid M},
+  publisher={Kluwer Academic Publishers},
+  year={2002}
+}
+
+@article{simpson1949measurement,
+  title={Measurement of diversity},
+  author={Simpson, Edward Hugh},
+  journal={Nature},
+  volume={163},
+  pages={688},
+  year={1949},
+  publisher={Nature Publishing Group},
+  url={https://doi.org/10.1038/163688a0}
+}
+
+@article{tarabichi2013systems,
+  title={Systems biology of cancer: entropy, disorder, and selection-driven evolution to independence, invasion and 'swarm intelligence'},
+  author={Tarabichi, Marta and Antoniou, Anna and Saiselet, Marjorie and Pita, Jos{\'e} M and Andry, G{\'e}rard and Dumont, Jos{\'e} E and others},
+  journal={Cancer Metastasis Reviews},
+  volume={32},
+  pages={403--421},
+  year={2013},
+  publisher={Springer},
+  url={https://doi.org/10.1007/s10555-013-9431-y}
+}
+
+@article{tsallis2017foundations,
+  title={On the foundations of statistical mechanics},
+  author={Tsallis, Constantino},
+  journal={European Physical Journal Special Topics},
+  volume={226},
+  pages={1433--1443},
+  year={2017},
+  publisher={Springer},
+  url={https://doi.org/10.1140/epjst/e2016-60252-2}
+}
+
+@article{van2014renyi,
+  title={Rényi Divergence and Kullback-Leibler Divergence},
+  author={Van Erven, Tim and Harremoes, Peter},
+  journal={IEEE Transactions on Information Theory},
+  volume={60},
+  number={7},
+  pages={3797--3820},
+  year={2014},
+  publisher={IEEE},
+  url={https://doi.org/10.1109/TIT.2014.2320500}
+}
+
+@article{yulmetyev2004dynamical,
+  title={Dynamical Shannon Entropy and Information},
+  author={Yulmetyev, Rustam M and Emelyanova, Natalya A and Gafarov, Farid M},
+  journal={Physica A},
+  volume={341},
+  pages={649--676},
+  year={2004},
+  publisher={Elsevier},
+  url={https://doi.org/10.1016/j.physa.2004.03.094}
+}
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
new file mode 100644
index 00000000000000..9d5dc9600e3a11
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
@@ -0,0 +1,358 @@
+---
+layout: tutorial_hands_on
+
+title: "Transcriptome Complexity Analysis with TSENAT"
+subtopic: information-theory
+priority: 10
+
+tags:
+    - rna-seq
+    - isoform
+    - entropy
+    - transcriptomics
+    - complexity
+    - information-theory
+
+level: Advanced
+time_estimation: 6h
+
+questions:
+    - What is Tsallis entropy and how does it relate to isoform complexity?
+    - How can I quantify transcriptome complexity changes between conditions?
+    - What are scale-dependent diversity measures and why do they matter?
+    - How do I detect isoform switching at different scales?
+    - What's the difference between Shannon entropy and Tsallis entropy?
+
+objectives:
+    - Understand the mathematical foundations of Tsallis entropy
+    - Explain the role of the entropic index (q) as a sensitivity parameter
+    - Describe the concept of scale-dependent isoform complexity
+    - Interpret q-curves and identify complexity patterns
+    - Understand the advantages of entropy-based analysis over traditional abundance measures
+    - Design and interpret scale-adaptive interaction models
+    - Identify scale-dependent isoform switching in RNA-seq data
+
+key_points:
+    - Tsallis entropy captures isoform complexity across multiple scales via the q parameter
+    - Traditional RNA-seq tools miss isoform reorganization without abundance changes
+    - Scale-adaptive interaction models detect condition-specific diversity patterns
+    - The q-spectrum reveals which biological scales are affected (rare vs abundant isoforms)
+    - Entropy-based approaches complement abundance-focused tools like DESeq2
+
+follow_up_training:
+    -
+        type: "internal"
+        topic_name: transcriptomics
+        tutorials:
+            - ref-based
+            - rna-seq-reads-to-counts
+
+contributions:
+  authorship:
+    - gallardoalba
+
+lang: en
+
+---
+
+This tutorial covers the theory and practical application of **TSENAT (Tsallis Entropy Analysis Toolbox)**, a  R package for quantifying and analyzing isoform-usage complexity in RNA-seq data using information-theoretic principles.
+
+
+> <agenda-title></agenda-title>
+>
+> In this tutorial, we will deal with:
+>
+> 1. TOC
+> {:toc}
+>
+{: .agenda}
+
+# Background: The Isoform Complexity Problem
+
+## Why Transcriptome Complexity Matters
+
+Standard RNA-seq analysis measures whether **transcript abundance** changes between conditions:
+- DESeq2 and edgeR ask: "Do genes produce more or less RNA?"
+- These tools focus on **total gene abundance**
+
+However, a critical complementary question remains largely unexplored:
+- **How does the diversity of isoforms change?**
+
+A gene may show **little change in total abundance** while dramatically **reshuffling its isoform repertoire**—a phenomenon often missed by traditional methods. This **isoform switching** reflects strategic shifts in protein function driven by splicing regulation, and represents genuine biological signal invisible to abundance-focused analyses.
+
+## Motivation: Evidence from the Literature
+
+### Cancer Heterogeneity and Network Entropy
+
+Recent evidence demonstrates that transcriptome complexity constitutes a measurable biological phenomenon:
+
+- **Tarabichi et al. (2013)**: Cancer cells show "increased entropy of signaling and gene interaction networks" as cancer progresses, contributing to cellular heterogeneity and evolvability
+- **Nijman (2020)**: Cancer-associated perturbations increase network entropy, driving phenotypic heterogeneity through disruption of gene regulatory networks
+- **Cao et al. (2017)**: Single-cell transcriptomics reveals that transcript-level complexity varies systematically across cell types and developmental states
+
+**The biological principle**: Changes in isoform organization—not just abundance—are valid biological signals worth measuring.
+
+## Abundance-Based Tools vs. Entropy-Based Tools
+
+| Aspect | DESeq2/edgeR | DRIMSeq | TSENAT |
+|--------|--------------|---------|--------|
+| **What it measures** | Total gene abundance | Differential transcript usage | Isoform heterogeneity/complexity |
+| **Detection capability** | Gene-level fold changes | Transcript switching | Reorganization of isoform landscape |
+| **A gene that doubles total count but reshuffles isoforms** | ✓ Detected as DE | Maybe detected | ✓ Detected as complexity change |
+| **A gene that maintains count but remixes isoforms** | ✗ Missed | Maybe detected | ✓ Detected as complexity change |
+| **Interpretability** | Fold-change magnitude | Proportional shift | Diversity pattern (rare vs abundant) |
+
+---
+
+# Theoretical Foundations: From Shannon to Tsallis Entropy
+
+## Historical Development
+
+### Claude Shannon (1948): Birth of Information Theory
+
+Claude Shannon's landmark 1948 work established that **information content could be quantified mathematically**. Shannon entropy measures the uncertainty when drawing a single observation from a probability distribution:
+
+$$H(p) = -\sum_{i=1}^{n} p_i \log_2(p_i)$$
+
+Where:
+- $p_i$ is the probability of outcome $i$
+- Higher entropy = more unpredictable outcome
+- In transcriptomics: higher entropy = more evenly distributed isoforms
+
+**Key insight**: If you draw a transcript from a gene's isoform pool, how predictable is the outcome?
+
+### Limitations of Shannon Entropy
+
+Shannon entropy treats all outcomes equally, regardless of frequency:
+- **Rare isoforms** and **common isoforms** receive identical weight
+- Cannot be tuned to emphasize one scale over another
+- Cannot separately examine complexity driven by rare variants vs. dominant isoforms
+
+### Generalized Entropy Families
+
+To address this limitation, mathematicians developed parametric entropy families:
+- **Rényi entropy** (1961): Introduced a tunable parameter α
+- **Tsallis entropy** (1988): Alternative generalization with parameter q
+- **Hill numbers** (2010): Modern ecological reinterpretation unifying all approaches
+
+**Remarkable fact**: Despite appearing different mathematically, Rényi and Tsallis entropies can be unified through generalized logarithmic and exponential functions—both answer the fundamental question: **What organizational scales matter?**
+
+## Mathematical Definition of Tsallis Entropy
+
+For a discrete probability vector **p** = (p₁, …, pₙ) representing isoform proportions within a gene, Tsallis entropy is defined as:
+
+$$S_q(p) = \frac{1 - \sum_{i=1}^{n} p_i^q}{q - 1}$$
+
+### Properties and Special Cases
+
+**The q parameter acts as a sensitivity dial:**
+
+- **Limit as q → 1**: Recovers Shannon entropy (balanced sensitivity)
+- **q = 0**: Richness (number of distinct isoforms present)
+- **q = 2**: Simpson index (emphasizes dominant isoforms)
+- **q < 1** (e.g., 0.5): Emphasizes **rare isoforms**
+- **q > 1** (e.g., 2.0): Emphasizes **abundant isoforms**
+
+### Biological Interpretation: Richness and Evenness
+
+**True diversity** decomposes into two independent components:
+
+1. **Richness**: How many distinct isoforms exist
+2. **Evenness**: How evenly transcripts are distributed
+
+Two genes can have identical Shannon entropy yet differ dramatically:
+
+| Scenario | Isoforms | Shannon Entropy | Interpretation |
+|----------|----------|-----------------|-----------------|
+| **Gene A** | 1 dominant + 99 rare | High (many isoforms contribute) | Heterogeneous: complexity across scales |
+| **Gene B** | 100 equally abundant | High (same total entropy) | Uniform: isoforms evenly distributed |
+
+By varying q, TSENAT reveals these differences: Gene A shows **high entropy at low q, declining at high q**, while Gene B shows **consistent entropy across all q-values**.
+
+---
+
+
+# Quality Control Approaches
+
+## M-Estimation for Sample Influence
+
+Before group-level comparisons, assess whether individual samples exert **disproportionate influence** on entropy estimates.
+
+### Leave-One-Out Influence Assessment
+
+1. **Robust M-estimation**: Iteratively re-weighted least squares with Huber loss
+2. **Design**: Quantifies how much sample removal affects location estimates
+3. **Independence from group assignment**: Sample-level QC at biological level
+4. **Interpretation**: Samples with high influence scores may indicate:
+   - Technical quality issues
+   - Subject-specific outliers
+   - Genuine biological heterogeneity
+
+### Flagging Criteria
+
+Samples warrant review if:
+- Distance from centroid > 1.5
+- Proportion of affected genes > 0.85
+- Outlier status flagged in QC analysis
+
+---
+
+# The Q-Spectrum: Visualization Framework
+
+## What is a Q-Curve?
+
+A **q-curve** plots Tsallis entropy across the entropic index spectrum from q=0 (rare-emphasizing) to q=∞ (abundant-emphasizing). 
+
+### Interpretation by Shape
+
+Three fundamental patterns emerge:
+
+1. **Flat curve (Balanced)**
+   - Entropy constant across q-values
+   - All isoforms shift proportionally
+   - Suggests systematic rebalancing
+   - Example: developmental transition
+
+2. **Declining curve (Rare-driven)**
+   - D(q=0.5) >> D(q=2.0)
+   - Low-abundance isoforms are condition-specific
+   - Treatment preferentially expresses rare transcripts
+   - Example: exploratory isoform expression
+
+3. **Rising curve (Abundant-driven)**
+   - D(q=0.5) << D(q=2.0)
+   - High-abundance isoforms shift between conditions
+   - Treatment remodels dominant transcript landscape
+   - Example: functional specialization
+
+### Multi-Scale Analysis
+
+Different biological processes prioritize different organizational scales:
+
+- **Isoform switching** (discrete change): Detected at specific q-values
+- **Wholesale reorganization** (widespread change): Detected at all q-values
+- **Cryptic isoform activation**: Revealed at low q (rare scale)
+- **Core architecture change**: Revealed at high q (abundant scale)
+
+---
+
+# Scale-Adaptive Interaction Models (SAIT)
+
+## The Core Question
+
+For each gene, we have a **q-curve**: entropy trajectory across entropic scales. We want to test: **Does this curve differ between groups?**
+
+Statistically, this is a **q × condition interaction**: the effect of condition on entropy depends on which q-value you examine.
+
+---
+
+# Transcript Switching Analysis
+
+## The Challenge: From Genes to Transcripts
+
+SAIT tells us **which genes** have scale-dependent complexity changes. But:
+- Which **specific transcripts** drive those changes?
+- Do the same transcripts switch across all scales?
+- Or do different regulatory mechanisms dominate at rare vs. abundant scales?
+
+## Jackknife Resampling Approach
+
+TSENAT uses **leave-one-out jackknife resampling** to identify robust transcript switches:
+
+### Delta Influence Metric
+
+For each transcript:
+- **Positive delta influence**: More important in control condition
+- **Negative delta influence**: More important in treatment condition
+- **Magnitude**: Robustness across bootstrap iterations
+- **Visualization**: Color intensity reflects consistency
+
+### Switching Classification
+
+1. **Consistent switching**: Same transcripts show strong delta influence across all q-values
+   - Robust, scale-independent splicing shift
+   - Suggests coordinated regulatory mechanism
+
+2. **Mixed/scale-dependent switching**: Different transcripts matter at different q-values
+   - Regulatory complexity varies by scale
+   - Different mechanisms for rare vs. abundant isoforms
+
+
+---
+
+# Tsallis Divergence: Effect Size Quantification
+
+## Definition
+
+For two probability distributions P (control) and Q (treatment) representing isoform proportions, **Tsallis divergence** measures information-theoretic distance:
+
+$$D_q(P||Q) = \frac{1 - \sum_i p_i^q \cdot q_i^{1-q}}{q - 1}$$
+
+Where:
+- D > 0 always (distance measure)
+- D = 0 only when distributions identical
+- Asymmetry: $D_q(P||Q) \ne D_q(Q||P)$ (information-theoretic property)
+
+## Biological Interpretation
+
+**Effect size threshold**: D > 0.1 indicates meaningful information-theoretic separation between conditions
+
+### Pattern Classification
+
+**Rare-driven complexity change**:
+- D(q=0.5) >> D(q=2.0)
+- Divergence concentrated at low q
+- Interpretation: Condition differences driven by rare transcripts
+
+**Abundant-driven complexity change**:
+- D(q=0.5) << D(q=2.0)
+- Divergence concentrated at high q
+- Interpretation: Condition differences in dominant isoforms
+
+**Balanced complexity change**:
+- D(q=0.5) ≈ D(q=2.0)
+- Divergence consistent across scales
+- Interpretation: All isoforms reorganize proportionally
+
+---
+
+# Method Validation: Non-Parametric Concordance
+
+## Why Two Statistical Methods?
+
+Statistical testing in transcriptomics faces a fundamental trade-off:
+- **Parametric methods (GAM)**: Maximum power when assumptions hold, but sensitive to violations
+- **Non-parametric methods (SRH)**: Robust to any distribution, but reduced power
+
+### Scheirer-Ray-Hare (SRH) Rank-Based Test
+
+**Advantages for entropy analysis**:
+
+1. **Robustness to distribution violations**: Operates on ranks (uniformly distributed), valid for ANY continuous distribution
+   - Non-normality: No effect on rank ordering
+   - Skewness: Rank transformation corrects automatically
+   - Bimodality: Inherently handled
+
+2. **Paired design for q-ordered structure**: 
+   - Treats q-measurements as within-subject repeated measures
+   - Captures sequential nature of q-spectrum
+   - AR(1) autocorrelation doesn't affect rank ordering (exchangeability property)
+
+3. **Interaction testing**:
+   - Two-way ANOVA on ranks
+   - Tests q × condition effects directly
+   - F-statistic follows F-distribution under null
+
+### Concordance Interpretation
+
+High agreement between GAM and SRH:
+- ✓ Robust discoveries across statistical frameworks
+- ✓ Results likely generalizable
+
+GAM-only discoveries:
+- Potential artifacts if GAM assumptions violated
+- Alternative: Trust only high-confidence genes significant in both methods
+
+SRH-only discoveries:
+- Rare; indicates distributional violations affecting GAM
+- May represent genuine signals from highly skewed distributions

From 87dc5b8969f02cb694a39ded67b5f2d31ff36a0b Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Mon, 20 Apr 2026 23:35:04 +0200
Subject: [PATCH 02/13] Fix issues

---
 .../tutorial.bib                              | 27 ++++++++++++-------
 .../tutorial.md                               | 14 +++++-----
 2 files changed, 26 insertions(+), 15 deletions(-)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.bib b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.bib
index e24318a8c641c7..9bd4d3091fbbb7 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.bib
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.bib
@@ -64,7 +64,8 @@ @article{benjamini1995controlling
   volume={57},
   number={1},
   pages={289--300},
-  year={1995}
+  year={1995},
+  doi={10.1111/j.2517-6161.1995.tb02031.x}
 }
 
 @article{cao2017comprehensive,
@@ -82,7 +83,8 @@ @book{chakraborty2019introductory
   title={Introductory Time Series Analysis},
   author={Chakraborty, Tapas R},
   publisher={Indian Statistical Institute, Kolkata},
-  year={2019}
+  year={2019},
+  isbn={9788183952934}
 }
 
 @article{chanda2020information,
@@ -114,7 +116,8 @@ @book{cover2006elements
   edition={2nd},
   author={Cover, Thomas M and Thomas, Joy A},
   publisher={Wiley-Interscience},
-  year={2006}
+  year={2006},
+  isbn={9780471241959}
 }
 
 @book{efron1993introduction,
@@ -122,7 +125,8 @@ @book{efron1993introduction
   edition={2nd},
   author={Efron, Bradley and Tibshirani, Robert J},
   publisher={Chapman and Hall},
-  year={1993}
+  year={1993},
+  isbn={0412042312}
 }
 
 @article{erhard2018improved,
@@ -133,7 +137,8 @@ @article{erhard2018improved
   number={12},
   pages={2096--2102},
   year={2018},
-  publisher={Oxford University Press}
+  publisher={Oxford University Press},
+  doi={10.1093/bioinformatics/bty056}
 }
 
 @article{ernst2004permutation,
@@ -143,7 +148,8 @@ @article{ernst2004permutation
   volume={19},
   number={4},
   pages={676--685},
-  year={2004}
+  year={2004},
+  doi={10.1214/088342304000000396}
 }
 
 @article{furuichi2006information,
@@ -253,7 +259,8 @@ @article{phipson2010permutation
   volume={9},
   number={1},
   pages={Article 39},
-  year={2010}
+  year={2010},
+  doi={10.2202/1544-6115.1585}
 }
 
 @article{ramirez2016determining,
@@ -310,14 +317,16 @@ @article{shannon1948mathematical
   volume={27},
   number={3--4},
   pages={379--423},
-  year={1948}
+  year={1948},
+  doi={10.1002/j.1538-7305.1948.tb01338.x}
 }
 
 @book{shiner2002entropy,
   title={Entropy and Entropy Generation: Fundamentals and Applications},
   author={Shiner, Jeffrey S and Emelyanova, Natalya A and Gafarov, Farid M},
   publisher={Kluwer Academic Publishers},
-  year={2002}
+  year={2002},
+  isbn={9781402006715}
 }
 
 @article{simpson1949measurement,
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
index 9d5dc9600e3a11..efb46ef782663e 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
@@ -135,7 +135,9 @@ To address this limitation, mathematicians developed parametric entropy families
 - **Tsallis entropy** (1988): Alternative generalization with parameter q
 - **Hill numbers** (2010): Modern ecological reinterpretation unifying all approaches
 
-**Remarkable fact**: Despite appearing different mathematically, Rényi and Tsallis entropies can be unified through generalized logarithmic and exponential functions—both answer the fundamental question: **What organizational scales matter?**
+#### Remarkable fact
+
+Despite appearing different mathematically, Rényi and Tsallis entropies can be unified through generalized logarithmic and exponential functions—both answer the fundamental question: **What organizational scales matter?**
 
 ## Mathematical Definition of Tsallis Entropy
 
@@ -145,7 +147,7 @@ $$S_q(p) = \frac{1 - \sum_{i=1}^{n} p_i^q}{q - 1}$$
 
 ### Properties and Special Cases
 
-**The q parameter acts as a sensitivity dial:**
+#### The q parameter acts as a sensitivity dial
 
 - **Limit as q → 1**: Recovers Shannon entropy (balanced sensitivity)
 - **q = 0**: Richness (number of distinct isoforms present)
@@ -299,17 +301,17 @@ Where:
 
 ### Pattern Classification
 
-**Rare-driven complexity change**:
+#### Rare-driven complexity change
 - D(q=0.5) >> D(q=2.0)
 - Divergence concentrated at low q
 - Interpretation: Condition differences driven by rare transcripts
 
-**Abundant-driven complexity change**:
+#### Abundant-driven complexity change
 - D(q=0.5) << D(q=2.0)
 - Divergence concentrated at high q
 - Interpretation: Condition differences in dominant isoforms
 
-**Balanced complexity change**:
+#### Balanced complexity change
 - D(q=0.5) ≈ D(q=2.0)
 - Divergence consistent across scales
 - Interpretation: All isoforms reorganize proportionally
@@ -326,7 +328,7 @@ Statistical testing in transcriptomics faces a fundamental trade-off:
 
 ### Scheirer-Ray-Hare (SRH) Rank-Based Test
 
-**Advantages for entropy analysis**:
+#### Advantages for entropy analysis
 
 1. **Robustness to distribution violations**: Operates on ranks (uniformly distributed), valid for ANY continuous distribution
    - Non-normality: No effect on rank ordering

From 507ab8594fc8178cfcc3c176488eaf7e6c77ca42 Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Mon, 20 Apr 2026 23:44:37 +0200
Subject: [PATCH 03/13] Remove subsection

---
 .../tutorials/transcriptome_complexity_tsenat/tutorial.md       | 2 --
 1 file changed, 2 deletions(-)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
index efb46ef782663e..883ecf6ca01464 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
@@ -2,8 +2,6 @@
 layout: tutorial_hands_on
 
 title: "Transcriptome Complexity Analysis with TSENAT"
-subtopic: information-theory
-priority: 10
 
 tags:
     - rna-seq

From e42f278a5cfa800beee97decd72ffbaeea76cdf0 Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Wed, 1 Jul 2026 00:35:33 +0200
Subject: [PATCH 04/13] Add draft slides

---
 .../slides.html                               | 36 +++++++++++++++++++
 1 file changed, 36 insertions(+)
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
new file mode 100644
index 00000000000000..a7945daed8d00e
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
@@ -0,0 +1,36 @@
+---
+layout: tutorial_slides
+logo: "GTN"
+
+title: "Isoform complexity analysis, or why dialectical logic works."
+contributions:
+  authorship:
+    - gallardoalba
+
+priority: 4
+
+video: no
+zenodo_link: ""
+
+
+---
+
+## Index of contents
+
+1. Introduction
+
+2. Galaxy workflow
+
+---
+
+## Introduction
+
+  - What is Tsallis entropy and how does it relate to isoform complexity?
+  
+  - What are scale-dependent diversity measures and why do they matter?
+  
+  - How do I detect isoform switching at different scales?
+
+  - Why dialectical logic works
+
+---

From 6b3b4e5849f0d2a44aba58b6225a9a0b9f9cb27b Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Wed, 1 Jul 2026 01:07:31 +0200
Subject: [PATCH 05/13] Update slides

---
 .../tutorials/transcriptome_complexity_tsenat/slides.html  | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
index a7945daed8d00e..18127066240389 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
@@ -34,3 +34,10 @@
   - Why dialectical logic works
 
 ---
+
+Ideas of order are profoundly ideological, so the description of evolution as producing order is necessarily an ideological one.
+
+.footnote[Source: [The Dialectical Biologist, p.53](https://www.hup.harvard.edu/books/9780674202832)]
+
+---
+

From 339aa7a681e1be288415375990806d452a5c7427 Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Wed, 1 Jul 2026 01:12:59 +0200
Subject: [PATCH 06/13] Fix citations

---
 .../transcriptome_complexity_tsenat/tutorial.md      | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
index 883ecf6ca01464..cea134a0a4b83f 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/tutorial.md
@@ -84,9 +84,9 @@ A gene may show **little change in total abundance** while dramatically **reshuf
 
 Recent evidence demonstrates that transcriptome complexity constitutes a measurable biological phenomenon:
 
-- **Tarabichi et al. (2013)**: Cancer cells show "increased entropy of signaling and gene interaction networks" as cancer progresses, contributing to cellular heterogeneity and evolvability
-- **Nijman (2020)**: Cancer-associated perturbations increase network entropy, driving phenotypic heterogeneity through disruption of gene regulatory networks
-- **Cao et al. (2017)**: Single-cell transcriptomics reveals that transcript-level complexity varies systematically across cell types and developmental states
+- **Tarabichi et al. (2013)**: {% cite tarabichi2013systems %} Cancer cells show "increased entropy of signaling and gene interaction networks" as cancer progresses, contributing to cellular heterogeneity and evolvability
+- **Nijman (2020)**: {% cite nijman2020perturbation %} Cancer-associated perturbations increase network entropy, driving phenotypic heterogeneity through disruption of gene regulatory networks
+- **Cao et al. (2017)**: {% cite cao2017comprehensive %} Single-cell transcriptomics reveals that transcript-level complexity varies systematically across cell types and developmental states
 
 **The biological principle**: Changes in isoform organization—not just abundance—are valid biological signals worth measuring.
 
@@ -129,9 +129,9 @@ Shannon entropy treats all outcomes equally, regardless of frequency:
 ### Generalized Entropy Families
 
 To address this limitation, mathematicians developed parametric entropy families:
-- **Rényi entropy** (1961): Introduced a tunable parameter α
-- **Tsallis entropy** (1988): Alternative generalization with parameter q
-- **Hill numbers** (2010): Modern ecological reinterpretation unifying all approaches
+- **Rényi entropy** (1961): Introduced a tunable parameter α {% cite van2014renyi %}
+- **Tsallis entropy** (1988): Alternative generalization with parameter q {% cite tsallis2017foundations %}
+- **Hill numbers** (2010): Modern ecological reinterpretation unifying all approaches {% cite chao2010phylogenetic %}
 
 #### Remarkable fact
 

From df4a23587bd98e3ed7675ce7b6d686a77f834fb6 Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Wed, 1 Jul 2026 01:26:53 +0200
Subject: [PATCH 07/13] slides.html

---
 .../transcriptome_complexity_tsenat/slides.html          | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
index 18127066240389..ed387467bddb1a 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
@@ -21,6 +21,15 @@
 
 2. Galaxy workflow
 
+---
+
+## Introduction
+
+To do science is to be a social actor engaged, whether one likes it or not, in political activity. The denial of the interpenetration of the scientific and the social is itself a political act [...]
+
+.footnote[Source: [The Dialectical Biologist, p.19](https://www.hup.harvard.edu/books/9780674202832)]
+
+
 ---
 
 ## Introduction

From 2877038b4cdd2996a74c0f11b35c478e4ee5a46c Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Tue, 7 Jul 2026 20:48:13 +0200
Subject: [PATCH 08/13] Update slides

---
 .../slides.html                               | 63 +++++++++++++++----
 1 file changed, 51 insertions(+), 12 deletions(-)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
index ed387467bddb1a..050e8c8e0519c0 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
@@ -23,30 +23,69 @@
 
 ---
 
-## Introduction
+# Introduction
 
-To do science is to be a social actor engaged, whether one likes it or not, in political activity. The denial of the interpenetration of the scientific and the social is itself a political act [...]
+> To do science is to be a social actor engaged, whether one likes it or not, in political activity. The denial of the interpenetration of the scientific and the social is itself a political act, giving support to social structures that hide behind scientific objectivity to perpetuate dependency, exploitation, racism, elitism, colonialism.
+> 
+> .footnote[Source: The Dialectical Biologist, Introduction](https://www.hup.harvard.edu/books/9780674202832)
 
-.footnote[Source: [The Dialectical Biologist, p.19](https://www.hup.harvard.edu/books/9780674202832)]
+---
+
+# Introduction
 
+     
+- The problem of measuring biological complexity
+- Parts, wholes, and dialectical relationships
+- Why reductionism fails for isoform complexity
+- Tsallis entropy: a dialectical measure
 
 ---
 
-## Introduction
+## The problem of measuring biological complexity
+
+- Reductionist approach: Count isoforms, compute ratios, test differences one by one
+- Problem: This treats each isoform as an independent entity with intrinsic properties
+- Cartesian assumption: "The part is ontologically prior to the whole" (Levins & Lewontin)
+
+     
+>  Cartesian reductionism is sometimes spoken of as the "Cartesian method"... But Cartesianism is more than simply a method of investigation; it is a commitment to how things really are.
+>
+> .footnote[Source: The Dialectical Biologist, Introduction]
+
+--- 
+
+## Parts, wholes, and dialectical relationships
 
-  - What is Tsallis entropy and how does it relate to isoform complexity?
-  
-  - What are scale-dependent diversity measures and why do they matter?
-  
-  - How do I detect isoform switching at different scales?
+> "Part" and "whole" have a special relationship to each other, in that one cannot exist without the other, any more than "up" can exist without "down." What constitutes the parts is defined by the whole that is being considered. Moreover, parts acquire properties by virtue of being parts of a particular whole, properties they do not have in isolation.
+>
+> .footnote[Source: The Dialectical Biologist, Introduction](https://www.hup.harvard.edu/books/9780674202832)
 
-  - Why dialectical logic works
+Key insight: Isoform diversity is not the sum of individual isoform abundances—it is an emergent property of the transcript distribution.
 
 ---
 
-Ideas of order are profoundly ideological, so the description of evolution as producing order is necessarily an ideological one.
+## Why reductionism fails for isoform complexity
 
-.footnote[Source: [The Dialectical Biologist, p.53](https://www.hup.harvard.edu/books/9780674202832)]
+- A gene with 5 isoforms where one dominates (99% abundance) has low complexity
+- A gene with 5 isoforms at equal abundance (20% each) has high complexity
+- Same number of parts, completely different wholes
+
+> It is not that the whole is more than the sum of its parts, but that the parts acquire new properties. But as the parts acquire properties by being together, they impart to the whole new properties, which are reflected in changes in the parts, and so on.
+> 
+> .footnote[Source: The Dialectical Biologist, Introduction](https://www.hup.harvard.edu/books/9780674202832)
 
 ---
 
+## Tsallis entropy: a dialectical measure
+
+Tsallis entropy captures diversity as a function of a single parameter q:
+
+$$S_q = \frac{1 - \sum_{i=1}^N p_i^q}{q - 1}$$
+
+- q → 0: Emphasizes rare isoforms (richness-dominated)
+- q = 1: Shannon entropy (standard measure)
+- q → 2+: Emphasizes abundant isoforms (evenness-dominated)
+
+Dialectical aspect: No single q-value reveals "the truth"—each scale reveals different properties of the whole.
+
+---
\ No newline at end of file

From d09965760977acbdedb69c2a18bc16bcc1aa6d2e Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Tue, 7 Jul 2026 20:49:03 +0200
Subject: [PATCH 09/13] Update slides

---
 .../tutorials/transcriptome_complexity_tsenat/slides.html | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
index 050e8c8e0519c0..63d7f3f5ae58d6 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
@@ -2,7 +2,7 @@
 layout: tutorial_slides
 logo: "GTN"
 
-title: "Isoform complexity analysis, or why dialectical logic works."
+title: "Isoform complexity analysis"
 contributions:
   authorship:
     - gallardoalba
@@ -23,8 +23,6 @@
 
 ---
 
-# Introduction
-
 > To do science is to be a social actor engaged, whether one likes it or not, in political activity. The denial of the interpenetration of the scientific and the social is itself a political act, giving support to social structures that hide behind scientific objectivity to perpetuate dependency, exploitation, racism, elitism, colonialism.
 > 
 > .footnote[Source: The Dialectical Biologist, Introduction](https://www.hup.harvard.edu/books/9780674202832)
@@ -33,7 +31,6 @@
 
 # Introduction
 
-     
 - The problem of measuring biological complexity
 - Parts, wholes, and dialectical relationships
 - Why reductionism fails for isoform complexity
@@ -47,10 +44,9 @@
 - Problem: This treats each isoform as an independent entity with intrinsic properties
 - Cartesian assumption: "The part is ontologically prior to the whole" (Levins & Lewontin)
 
-     
 >  Cartesian reductionism is sometimes spoken of as the "Cartesian method"... But Cartesianism is more than simply a method of investigation; it is a commitment to how things really are.
 >
-> .footnote[Source: The Dialectical Biologist, Introduction]
+> .footnote[Source: The Dialectical Biologist, Introduction](https://www.hup.harvard.edu/books/9780674202832)
 
 --- 
 

From 0ace751148dd8e5c36baf260224892b63eada0ad Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Tue, 7 Jul 2026 20:50:42 +0200
Subject: [PATCH 10/13] Update slides

---
 .../tutorials/transcriptome_complexity_tsenat/slides.html       | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
index 63d7f3f5ae58d6..c250840942e0f8 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
@@ -23,7 +23,7 @@
 
 ---
 
-> To do science is to be a social actor engaged, whether one likes it or not, in political activity. The denial of the interpenetration of the scientific and the social is itself a political act, giving support to social structures that hide behind scientific objectivity to perpetuate dependency, exploitation, racism, elitism, colonialism.
+> To do science is to be a social actor engaged, whether one likes it or not, in political activity. The denial of the interpenetration of the scientific and the social is itself a political act [...] </div>
 > 
 > .footnote[Source: The Dialectical Biologist, Introduction](https://www.hup.harvard.edu/books/9780674202832)
 

From 05bd201c7865ea89287574f7780c4ad6be94c7e1 Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Tue, 7 Jul 2026 20:53:15 +0200
Subject: [PATCH 11/13] Update slides

---
 .../tutorials/transcriptome_complexity_tsenat/slides.html       | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
index c250840942e0f8..d6434fc3421eb2 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
@@ -23,7 +23,7 @@
 
 ---
 
-> To do science is to be a social actor engaged, whether one likes it or not, in political activity. The denial of the interpenetration of the scientific and the social is itself a political act [...] </div>
+> To do science is to be a social actor engaged, whether one likes it or not, in political activity. The denial of the interpenetration of the scientific and the social is itself a political act [...]
 > 
 > .footnote[Source: The Dialectical Biologist, Introduction](https://www.hup.harvard.edu/books/9780674202832)
 

From 247deb3d6959f84843a4e3160acea914d24f6b9b Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Tue, 7 Jul 2026 20:54:45 +0200
Subject: [PATCH 12/13] Update slides

---
 .../tutorials/transcriptome_complexity_tsenat/slides.html       | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
index d6434fc3421eb2..1b6c05abba91ff 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/slides.html
@@ -56,7 +56,7 @@
 >
 > .footnote[Source: The Dialectical Biologist, Introduction](https://www.hup.harvard.edu/books/9780674202832)
 
-Key insight: Isoform diversity is not the sum of individual isoform abundances—it is an emergent property of the transcript distribution.
+Key insight: Isoform complexity is not the sum of individual isoform abundances—it is an emergent property of the transcript distribution.
 
 ---
 

From 9733f571c158699f036316ebf8a53b7498ece7a4 Mon Sep 17 00:00:00 2001
From: gallardoalba <gallardoalbac@gmail.com>
Date: Tue, 7 Jul 2026 20:59:20 +0200
Subject: [PATCH 13/13] Fix FAQs format

---
 .../faqs/complexity_vs_deseq2.md              |  14 ++
 .../faqs/effect_size_reporting.md             |  16 ++
 .../faqs/erratic_q_curves.md                  |  13 ++
 .../faqs/index.md                             | 143 +-----------------
 .../faqs/long_reads.md                        |  13 ++
 .../faqs/na_nan_results.md                    |  14 ++
 .../faqs/no_significant_genes.md              |  13 ++
 .../faqs/q_condition_interaction.md           |  12 ++
 .../faqs/q_spectrum_choice.md                 |  12 ++
 .../faqs/read_depth_entropy.md                |  13 ++
 .../faqs/sample_size.md                       |  12 ++
 .../faqs/transcript_counts.md                 |  14 ++
 .../faqs/transcript_switching_types.md        |  15 ++
 .../faqs/tsallis_divergence_vs_foldchange.md  |  18 +++
 .../faqs/tsallis_vs_shannon.md                |  14 ++
 15 files changed, 198 insertions(+), 138 deletions(-)
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/complexity_vs_deseq2.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/effect_size_reporting.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/erratic_q_curves.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/long_reads.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/na_nan_results.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/no_significant_genes.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/q_condition_interaction.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/q_spectrum_choice.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/read_depth_entropy.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/sample_size.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/transcript_counts.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/transcript_switching_types.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/tsallis_divergence_vs_foldchange.md
 create mode 100644 topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/tsallis_vs_shannon.md

diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/complexity_vs_deseq2.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/complexity_vs_deseq2.md
new file mode 100644
index 00000000000000..ab6c21333852cb
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/complexity_vs_deseq2.md
@@ -0,0 +1,14 @@
+---
+title: Why is transcriptome complexity important if DESeq2 already detects changes?
+area: conceptual
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+DESeq2 and edgeR measure **total gene abundance** changes. They miss situations where:
+- A gene maintains constant total expression but **reorganizes its isoforms**
+- Rare isoforms become more common (or vice versa) without affecting overall counts
+- The **diversity** of the isoform landscape changes
+
+These are valid biological signals (isoform switching via splicing regulation) that complement abundance-based methods.
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/effect_size_reporting.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/effect_size_reporting.md
new file mode 100644
index 00000000000000..8b2142e6bc74b6
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/effect_size_reporting.md
@@ -0,0 +1,16 @@
+---
+title: A gene shows significant q×condition interaction but small effect size—should I report it?
+area: interpretation
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+Guidelines for meaningful results:
+- **Effect size (Tsallis divergence) D > 0.1**: ~10% information divergence (notable)
+- **D > 0.2**: Substantial divergence (recommended threshold)
+- **Small D with p < 0.05**: May reflect statistical noise; prioritize high-effect genes
+
+Report both for completeness:
+- High-effect genes (D > 0.2): Strong biological signal
+- Borderline genes (D ~ 0.1, p < 0.05): Mention as candidates for validation
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/erratic_q_curves.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/erratic_q_curves.md
new file mode 100644
index 00000000000000..06bd7731351bc8
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/erratic_q_curves.md
@@ -0,0 +1,13 @@
+---
+title: Q-curves show erratic patterns. Is this normal?
+area: troubleshooting
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+**Expected**: Small fluctuations in entropy across q-values
+**Concerning**: Large spikes or reversals
+- May indicate normalization issues
+- Check: Are housekeeping genes (GAPDH, ACTB) showing balanced patterns?
+- If not: Revisit pseudocount and filtering parameters
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/index.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/index.md
index 821d165ce527a4..fde41fcf19a8d4 100644
--- a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/index.md
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/index.md
@@ -1,138 +1,5 @@
-# Frequently Asked Questions about TSENAT and Transcriptome Complexity Analysis
-
-## Conceptual Questions
-
-### What's the difference between Tsallis entropy and Shannon entropy?
-
-**Shannon entropy** (q=1 in Tsallis framework) treats all isoforms equally. **Tsallis entropy** adds a parameter `q` that acts as a sensitivity dial:
-- When q < 1: emphasizes **rare isoforms**
-- When q = 1: recovers Shannon entropy (balanced)
-- When q > 1: emphasizes **abundant isoforms**
-
-This allows you to examine complexity at different biological scales in a single analysis.
-
-### Why is transcriptome complexity important if DESeq2 already detects changes?
-
-DESeq2 and edgeR measure **total gene abundance** changes. They miss situations where:
-- A gene maintains constant total expression but **reorganizes its isoforms**
-- Rare isoforms become more common (or vice versa) without affecting overall counts
-- The **diversity** of the isoform landscape changes
-
-These are valid biological signals (isoform switching via splicing regulation) that complement abundance-based methods.
-
-## Technical Questions
-
-### What sample size do I need?
-
-TSENAT assumes **≥6-8 samples per group** for paired/longitudinal designs:
-- Smaller samples are underpowered to detect entropy shifts
-- Larger samples increase power to detect subtle complexity changes
-- Biological replicates are essential (not technical replicates)
-
-### How do I choose the q-spectrum (range and granularity)?
-
-**Standard configuration**: q = 0 to 2.0 with Δq = 0.05 (41 q-values)
-- Range: 0-2 covers extreme rare-emphasis to abundant-emphasis
-- Granularity: Δq=0.05 provides good resolution without computational burden
-- For visualization: Can downsample to q = (0, 0.5, 1, 1.5, 2) for quick exploration
-
-### Does read depth affect entropy estimates?
-
-Yes. TSENAT uses **automatic pseudocount selection** to handle this:
-- Pseudocounts scale with library size
-- Ensures entropy estimates are comparable across samples with different sequencing depths
-- Breaks ties in rank-based validation methods
-- Standard practice in RNA-seq analysis
-
-### What's the difference between "consistent" vs. "scale-dependent" transcript switching?
-
-- **Consistent switching**: Same transcripts show high delta-influence values across all q-values
-  - Robust, scale-independent splicing shift
-  - Suggests coordinated regulatory mechanism
-  
-- **Scale-dependent switching**: Different transcripts matter at different q-values
-  - Regulatory complexity varies by scale
-  - Different mechanisms for rare vs. abundant isoforms
-
-## Methodological Questions
-
-### What does "q × condition interaction" mean?
-
-It means the effect of **condition on entropy depends on which q-value** you examine:
-- At q=0.5, one condition might have higher entropy
-- At q=2, the pattern might reverse
-- This reveals **scale-dependent** biological signals
-
-### What's the advantage of Tsallis divergence over fold-change?
-
-**Fold-change**:
-- Simple 1:1 ratio between conditions
-- Misses complexity patterns
-- Unbounded (can be arbitrarily large)
-
-**Tsallis divergence**:
-- Information-theoretic distance between distributions
-- Captures pattern type (rare-driven, balanced, abundant-driven)
-- Symmetric [0] to asymmetric, bounded interpretability
-- Directly quantifies isoform complexity differences
-
-## Practical Workflow Questions
-
-### Where do I get transcript-level counts for this analysis?
-
-TSENAT expects **SALMON or Kallisto output** including:
-- NumReads: raw fragment counts per transcript
-- TPM: length- and library-normalized estimates
-- EffectiveLength: read-length corrected transcript length
-
-These are standard outputs from pseudoalignment-based quantification.
-
-
-## Interpretation Questions
-
-### A gene shows significant q×condition interaction but small effect size—should I report it?
-
-Guidelines for meaningful results:
-- **Effect size (Tsallis divergence) D > 0.1**: ~10% information divergence (notable)
-- **D > 0.2**: Substantial divergence (recommended threshold)
-- **Small D with p < 0.05**: May reflect statistical noise; prioritize high-effect genes
-
-Report both for completeness:
-- High-effect genes (D > 0.2): Strong biological signal
-- Borderline genes (D ~ 0.1, p < 0.05): Mention as candidates for validation
-
-
-## Troubleshooting
-
-### My results show no significant genes. What could be wrong?
-
-1. **Sample size too small**: Need ≥6-8 per group (check design)
-2. **Entropy differences too subtle**: Try visualizing q-curves manually
-3. **Filtering too stringent**: Relaxing filters may reveal additional patterns
-4. **Method choice**: Try alternative statistical method (GAM vs. LMM vs. SRH)
-5. **Data quality**: Run M-estimation QC; check for outlier samples
-
-### Some q-values give NA or NaN results. Why?
-
-Possible causes:
-1. **Zero counts**: Pseudocount selection should handle; check filtering
-2. **Numerical instability**: Very small divergence values can underflow
-3. **Method-specific**: SRH (rank test) may have issues with many tied values
-
-Solution: Ensure pseudocount regularization is enabled; try alternative statistical method.
-
-### Q-curves show erratic patterns. Is this normal?
-
-**Expected**: Small fluctuations in entropy across q-values
-**Concerning**: Large spikes or reversals
-- May indicate normalization issues
-- Check: Are housekeeping genes (GAPDH, ACTB) showing balanced patterns?
-- If not: Revisit pseudocount and filtering parameters
-
-### Can I use TSENAT with very long reads (PacBio, Nanopore)?
-
-Conceptually yes, but:
-- TSENAT expects pseudoalignment-based quantification (SALMON, Kallisto)
-- Long reads typically use alignment-based tools (Minimap2, etc.)
-- Workflow would need adaptation to parse alignment-based counts
-- Overall approach (entropy framework) remains valid
+---
+layout: faq-page
+area_order: [conceptual, technical, methodological, workflow, interpretation, troubleshooting]
+title: TSENAT FAQs
+---
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/long_reads.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/long_reads.md
new file mode 100644
index 00000000000000..27c40e6663ea2f
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/long_reads.md
@@ -0,0 +1,13 @@
+---
+title: Can I use TSENAT with very long reads (PacBio, Nanopore)?
+area: troubleshooting
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+Conceptually yes, but:
+- TSENAT expects pseudoalignment-based quantification (SALMON, Kallisto)
+- Long reads typically use alignment-based tools (Minimap2, etc.)
+- Workflow would need adaptation to parse alignment-based counts
+- Overall approach (entropy framework) remains valid
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/na_nan_results.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/na_nan_results.md
new file mode 100644
index 00000000000000..07062eba33f2eb
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/na_nan_results.md
@@ -0,0 +1,14 @@
+---
+title: Some q-values give NA or NaN results. Why?
+area: troubleshooting
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+Possible causes:
+1. **Zero counts**: Pseudocount selection should handle; check filtering
+2. **Numerical instability**: Very small divergence values can underflow
+3. **Method-specific**: SRH (rank test) may have issues with many tied values
+
+Solution: Ensure pseudocount regularization is enabled; try alternative statistical method.
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/no_significant_genes.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/no_significant_genes.md
new file mode 100644
index 00000000000000..2ae152ed5610c6
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/no_significant_genes.md
@@ -0,0 +1,13 @@
+---
+title: My results show no significant genes. What could be wrong?
+area: troubleshooting
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+1. **Sample size too small**: Need ≥6-8 per group (check design)
+2. **Entropy differences too subtle**: Try visualizing q-curves manually
+3. **Filtering too stringent**: Relaxing filters may reveal additional patterns
+4. **Method choice**: Try alternative statistical method (GAM vs. LMM vs. SRH)
+5. **Data quality**: Run M-estimation QC; check for outlier samples
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/q_condition_interaction.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/q_condition_interaction.md
new file mode 100644
index 00000000000000..78a25fce8f4839
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/q_condition_interaction.md
@@ -0,0 +1,12 @@
+---
+title: What does "q × condition interaction" mean?
+area: methodological
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+It means the effect of **condition on entropy depends on which q-value** you examine:
+- At q=0.5, one condition might have higher entropy
+- At q=2, the pattern might reverse
+- This reveals **scale-dependent** biological signals
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/q_spectrum_choice.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/q_spectrum_choice.md
new file mode 100644
index 00000000000000..0ce452cdf7dbd9
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/q_spectrum_choice.md
@@ -0,0 +1,12 @@
+---
+title: How do I choose the q-spectrum (range and granularity)?
+area: technical
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+**Standard configuration**: q = 0 to 2.0 with Δq = 0.05 (41 q-values)
+- Range: 0-2 covers extreme rare-emphasis to abundant-emphasis
+- Granularity: Δq=0.05 provides good resolution without computational burden
+- For visualization: Can downsample to q = (0, 0.5, 1, 1.5, 2) for quick exploration
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/read_depth_entropy.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/read_depth_entropy.md
new file mode 100644
index 00000000000000..15ff56d1149e21
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/read_depth_entropy.md
@@ -0,0 +1,13 @@
+---
+title: Does read depth affect entropy estimates?
+area: technical
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+Yes. TSENAT uses **automatic pseudocount selection** to handle this:
+- Pseudocounts scale with library size
+- Ensures entropy estimates are comparable across samples with different sequencing depths
+- Breaks ties in rank-based validation methods
+- Standard practice in RNA-seq analysis
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/sample_size.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/sample_size.md
new file mode 100644
index 00000000000000..6a1c563de8e446
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/sample_size.md
@@ -0,0 +1,12 @@
+---
+title: What sample size do I need?
+area: technical
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+TSENAT assumes **≥6-8 samples per group** for paired/longitudinal designs:
+- Smaller samples are underpowered to detect entropy shifts
+- Larger samples increase power to detect subtle complexity changes
+- Biological replicates are essential (not technical replicates)
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/transcript_counts.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/transcript_counts.md
new file mode 100644
index 00000000000000..1f5e5cea5c6abf
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/transcript_counts.md
@@ -0,0 +1,14 @@
+---
+title: Where do I get transcript-level counts for this analysis?
+area: workflow
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+TSENAT expects **SALMON or Kallisto output** including:
+- NumReads: raw fragment counts per transcript
+- TPM: length- and library-normalized estimates
+- EffectiveLength: read-length corrected transcript length
+
+These are standard outputs from pseudoalignment-based quantification.
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/transcript_switching_types.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/transcript_switching_types.md
new file mode 100644
index 00000000000000..128dced9b788d2
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/transcript_switching_types.md
@@ -0,0 +1,15 @@
+---
+title: What's the difference between "consistent" vs. "scale-dependent" transcript switching?
+area: technical
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+- **Consistent switching**: Same transcripts show high delta-influence values across all q-values
+  - Robust, scale-independent splicing shift
+  - Suggests coordinated regulatory mechanism
+  
+- **Scale-dependent switching**: Different transcripts matter at different q-values
+  - Regulatory complexity varies by scale
+  - Different mechanisms for rare vs. abundant isoforms
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/tsallis_divergence_vs_foldchange.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/tsallis_divergence_vs_foldchange.md
new file mode 100644
index 00000000000000..72cb16a5abbd5c
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/tsallis_divergence_vs_foldchange.md
@@ -0,0 +1,18 @@
+---
+title: What's the advantage of Tsallis divergence over fold-change?
+area: methodological
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+**Fold-change**:
+- Simple 1:1 ratio between conditions
+- Misses complexity patterns
+- Unbounded (can be arbitrarily large)
+
+**Tsallis divergence**:
+- Information-theoretic distance between distributions
+- Captures pattern type (rare-driven, balanced, abundant-driven)
+- Symmetric [0] to asymmetric, bounded interpretability
+- Directly quantifies isoform complexity differences
diff --git a/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/tsallis_vs_shannon.md b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/tsallis_vs_shannon.md
new file mode 100644
index 00000000000000..2ed36924ccedbd
--- /dev/null
+++ b/topics/transcriptomics/tutorials/transcriptome_complexity_tsenat/faqs/tsallis_vs_shannon.md
@@ -0,0 +1,14 @@
+---
+title: What's the difference between Tsallis entropy and Shannon entropy?
+area: conceptual
+layout: faq
+box_type: tip
+contributors: [gallardoalba]
+---
+
+**Shannon entropy** (q=1 in Tsallis framework) treats all isoforms equally. **Tsallis entropy** adds a parameter `q` that acts as a sensitivity dial:
+- When q < 1: emphasizes **rare isoforms**
+- When q = 1: recovers Shannon entropy (balanced)
+- When q > 1: emphasizes **abundant isoforms**
+
+This allows you to examine complexity at different biological scales in a single analysis.