Download raw tables, code snippets, and datasets from full texts to power analyses.
Press Enter ↵ to solve
Fuel Your Discoveries
"The scientist only imposes two things, namely truth and sincerity, imposes them upon himself and upon other scientists."
- Erwin Schrödinger
Quick Analysis Plan
Copied
Causality-first meta-analysis plan (Arabidopsis regeneration epigenetics)
Use the supplied Arabidopsis/regeneration-relevant datasets to (i) map chromatin-marker dynamics (DNA methylation, histone marks, 3D contacts, sRNAs) across perturbations (RdDM/ROS1, chromatin modifiers, TFs), then (ii) test marker→regeneration/identity causality using rigorous design logic (orthogonal perturbations + directionality checks), and (iii) synthesize a cross-study “mechanism scorecard” anchored by Tea-IR/Ea-IR immune looping , WIND1 acetylation switch during somatic embryogenesis , and histone variant & memory frameworks .
Long Analysis Plan
Meta-analysis: Plant regeneration epigenetic control — causality-first plan
Goal (what we will answer): Across Arabidopsis regeneration-relevant perturbations, determine which chromatin markers (DNA methylation, histone modifications like H3K27Ac / H3K4me3 / H3K27me3, 3D contact changes, and sRNA production) change in a directionally consistent way with regeneration/identity outcomes—then test whether those marker changes are plausibly causal using orthogonal perturbations and mechanism constraints.
1) Evidence map (VISUAL)
We anchor the meta-analysis on the supplied core mechanistic studies and their explicit data dependencies (e.g., marker assays, perturbation types, and organism/tissue contexts).
Study anchors used (for claims below):
2) Data acquisition & harmonization plan (what we will download + how we will align them)
3C/sRNA/BS-seq integration for chromatin topology + RdDM: Use public GEO resources listed for the Ea-IR work (e.g., GSE79780, GSE19694, GSE57191, GSE80744) and interpret alongside the paper’s locus architecture assumptions (Ea-IR between EFR and XI-k).
H3K27Ac + WIND1 binding + RNA reprogramming during somatic embryogenesis: Download ChIP-seq and RNA-seq data from GEO BioProjects referenced (GSE301977 ChIP-seq; GSE302400 RNA-seq).
Chromatin “resetting” and fate coupling for cell-cycle/PRC2 memory: Use the provided repository for the computational model and the paper’s described experimental logic (H3K27me3 at STM diluted by division).
Methylation-sensing circuit as a marker-stability control axis: Download methylation data from GEO accession GSE104240 for the ROS1 rheostat study.
Oxygen-priming and H3K4me3 dynamics: Use the stated Bioproject PRJNA1346290 for RNA-seq and the immunostaining/data described as mechanistic context (root tips, controlled O2).
2.2 Harmonize across marker layers (no mixed conclusions)
Coordinate systems: unify genome build references when possible (e.g., the Ea-IR paper references TAIR10 and describes older TAIR9 usage).
Marker quantification standardization: convert per-condition measures into comparable effect representations within each study (e.g., log2 fold-change for RNA, differential methylation summaries by context CG/CHG/CHH, ChIP peak intensity changes, and loop/opening metrics where provided).
Strict directionality rules for causality scoring: only compare direction within a study’s perturbation axis; cross-study “marker correlation” is treated as a hypothesis generator, not causality, unless multiple orthogonal perturbations converge.
Because these studies are rich in perturbations, we can build a marker causality score using rules that require more than “marker changes happened.” We will explicitly separate:
(i) co-occurrence (marker changes with phenotype) from
(ii) dependency (marker changes require a specific pathway) and
(iii) intervention-like logic (orthogonal perturbations change the same directionally consistent mechanism).
Important skepticism note: this “scorecard” is not a statistical meta-analysis yet—it is a structured decision rubric for what we will compute from the raw data. The numeric values above reflect only the presence of mechanistic dependency described in the supplied excerpts (e.g., WIND1 recruiting HDA9 + ADA2a-HAG1; ROS1 rheostat buffering; division diluting PRC2 H3K27me3 at STM), not effect sizes.
4) Quantitative meta-analysis outputs (what we will compute)
4.1 Marker→target gene directionality matrix
For each study, define “regeneration outcome” genes: e.g., WIND1 embryogenesis regulators (LEC2) and shoot identity genes (ANT, SPL9) described as repressed/activated with H3K27Ac changes.
For Ea-IR: define EFR-associated outcomes and Loop EFR vs Loop XI-k signatures; incorporate methylation contexts and RdDM dependence.
Counts used are explicitly stated for WIND1 direct binding (616 activated and 114 repressed) in the provided excerpt.
Build a graph where nodes are chromatin markers/pathways (RdDM/Ea-IR siRNA loop resetting; WIND1→HDA9/ADA2a-HAG1; PRC2-mediated H3K27me3 dilution by division; KDM7→H3K4me3 under hypoxia; HDA19→H3.3K27/K36ac; ROS1 methylation-sensing circuit; histone variant logic), and edges are perturbation→marker and marker→outcome relationships described in each study.
Network edges are conceptual mechanism linkages that mirror explicitly described study logic (e.g., WIND1 recruits HDA9 and ADA2a-HAG1 to drive H3K27Ac-linked transcriptional reprogramming).
5) Implementation blueprint (code-level steps, but no example datasets)
5.1 Pipeline structure
Ingest: pull raw reads (where available) + metadata (sample group, timepoints, perturbations) for the specified GEO/SRA accessions: Ea-IR GEO accessions (GSE79780, GSE19694, GSE57191, GSE80744) and WIND1 (GSE301977 ChIP-seq; GSE302400 RNA-seq) plus ROS1 inheritance (GSE104240).
Preprocess per modality:
RNA-seq: produce gene-level fold-changes for regeneration-relevant genes described in each study (e.g., WIND1 regulators).
ChIP-seq: call peaks/broad signals for histone marks used in each paper (e.g., H3K27Ac under WIND1).
WGBS/BS-seq: compute differential methylation summarized by context (CG/CHG/CHH where described).
3C-qPCR/contacts: extract loop-opening metrics as available in the study’s dataset outputs (or from reported values if raw not downloadable).
Within-study causality triads:
Triad A (WIND1): WIND1 perturbation → H3K27Ac change → fate-gene expression → somatic embryogenesis outcome; validate dependency on HDA9 and ADA2a-HAG1 modules using described inhibitor/mutant axes.
Triad C (cell-cycle): division rate/perturbation → STM H3K27me3 dynamics → STM expression and stem identity maintenance; quiescence irreversibly shifts to differentiation.
Cross-study synthesis: “mechanism congruence”
Compute whether marker changes are consistent with the direction expected by mechanism type (acetylation switch vs methylation rheostat vs 3D loop resetting).
Down-weight studies where the excerpt flags correlation-vs-causality uncertainty or confounding risks (e.g., aphid selection weak evidence and prior re-analysis about misassignment).
Correlation vs causation: marker changes may be downstream readouts of regeneration; we require orthogonal perturbations within a study (e.g., WIND1 recruitment modules; ROS1 rheostat disruption) before treating marker↔phenotype as causal.
Model/reporter artifacts: inducible overexpression and inhibitor use can introduce non-physiological effects; we will treat those results as conditional and keep “causal confidence” lower when only such perturbations exist.
Species/context overreach: most mechanistic data are Arabidopsis-centered; we will not claim universal principles outside the provided cross-species scope of reviews.
Confounding in multigenerational inheritance claims: aphid selection shows weak evidence and sensitivity to accession-treatment confounding; we will treat heritable epigenetic claims as high-uncertainty unless replication is strong and confounding controls are explicit.
Generalizability of immune epigenetics to regeneration: Ea-IR study is pathogen immunity-focused; we will only use it for mechanism-type insights (e.g., 3D loop + siRNA resetting) rather than asserting it equals regeneration biology.
7) Deliverables checklist (what the user will get as outputs)
Effect-direction tables (marker layer × target gene module × perturbation) for each anchored study.
Mechanism congruence score per marker type (acetylation-switch, methylation-rheostat, PRC2 dilution, oxygen-sensing, 3D loop+siRNA reset), with explicit “evidence type” tracking (dependency vs co-occurrence).
Network visualization mapping perturbations to marker layers to outcomes (provided above as a mechanism map scaffold).
Confidence report that states what would disprove the causal model (e.g., marker changes without phenotype, phenotype without marker dependency), aligned to each study’s own falsification logic where available in the excerpts.
Key disproof targets (operational):
If WIND1-mediated somatic embryogenesis can occur without the HDA9/ADA2a-HAG1-dependent H3K27Ac reprogramming described, the acetylation-switch causality would be weakened.
If Ea-IR absence does not alter Loop EFR repression, EFR basal levels, or pathogen resistance in the way described, the 3D loop + RdDM resetting mechanism would be undermined.
It is ingesting GEO raw data for WIND1 (GSE301977/302400) and Ea-IR (GSE79780/19694/57191/80744), then computes per-perturbation effect directions linking marker layers to fate-gene modules.
Get emailed when your analysis is done!
We'll email you the results when your analysis is finished.
Hypothesis Graveyard
A simple “one histone mark explains regeneration” model (e.g., H3K27Ac alone) is unlikely because the provided studies show distinct causal axes (deacetylation/acetylation recruitment in WIND1, PRC2/H3K27me3 dilution via division, KDM7/H3K4me3 oxygen sensing, ROS1 methylation rheostat stability).
A “transgenerational inheritance is the primary driver” hypothesis is weakened by the aphid-selection evidence showing weak heritable epigenetic differentiation and confounding sensitivity; within-individual or within-development memory appears more robust than strong cross-generation inheritance in these excerpts.
Quick Analysis Plan
Long Analysis Plan
Top Data Sources
ExportMCP
Keep Exploring
Analysis Wizard
Hypothesis Graveyard
Potential Experiments
Discussion
Fuel Your Discoveries
Ask a Follow-Up
Key Insight