---
title: "Calcifying Contra"
description: "and beyond"
categories: ["Transcriptomics"]
#citation:
date: 11-29-2025
image: http://gannet.fish.washington.edu/seashell/snaps/Monosnap_Image_2025-11-30_08-27-36.png # finding a good image
author:
- name: Steven Roberts
url:
orcid: 0000-0001-8302-1138
affiliation: Professor, UW - School of Aquatic and Fishery Sciences
affiliation-url: https://robertslab.info
#url: # self-defined
draft: false # setting this to `true` will prevent your post from appearing on your listing page until you're ready!
format:
html:
toc: true # ← enable TOC
toc-location: left # or: right, body
toc-depth: 3 # how many heading levels to show
code-fold: FALSE
code-tools: true
code-copy: true
highlight-style: github
code-overflow: wrap
#runtime: shiny
---
First off in time-series-calcification
https://github.com/urol-e5/timeseries-molecular-calcification/blob/main/M-multi-species/scripts/34-biomin-pathway-compare.rmd
 
 
# Contra
Then took onto ConTra... with new script
```
Duro:ConTra sr320$ python3 code/run_biomin_species_comparison.py ======================================================================
🧬 CROSS-SPECIES BIOMIN CONTEXT-DEPENDENT ANALYSIS
======================================================================
Workspace: /Users/sr320/GitHub/ConTra
Comparison output: /Users/sr320/GitHub/ConTra/output/biomin_comparison_20251129_144655
======================================================================
```
```
Using context_dependent_analysis_20251129_192544 for apul
Using context_dependent_analysis_20251129_162258 for peve
Using context_dependent_analysis_20251129_144655 for ptua
============================================================
LOADING SPECIES RESULTS
============================================================
📂 Loading A. pulchra...
✅ Loaded methylation_mirna_context.csv: 33270 rows
📂 Loading P. evermanni...
✅ Loaded methylation_mirna_context.csv: 27150 rows
📂 Loading P. tuahiniensis...
✅ Loaded methylation_mirna_context.csv: 51150 rows
============================================================
COMPARING METHYLATION-MIRNA CONTEXT ACROSS SPECIES
============================================================
📊 Summary:
Total unique OGs across all species: 611
OGs in 2+ species: 358
OGs in all 3 species: 100
✅ Saved comparison table: /Users/sr320/GitHub/ConTra/output/biomin_comparison_20251130_052506/cross_species_methylation_mirna_comparison.csv
🏆 Top 15 Conserved Context-Dependent OGs (in 2+ species):
og_id n_species mean_context_strength apul_context_strength peve_context_strength ptua_context_strength
OG_01155 3 1.310726 1.521005 0.996267 1.414906
OG_02619 3 1.295301 1.468612 1.024492 1.392800
OG_06119 3 1.236816 0.971755 1.252635 1.486059
OG_08793 3 1.236346 0.808764 1.450764 1.449511
OG_05366 3 1.199557 0.939276 1.174450 1.484946
OG_03291 3 1.199266 1.186752 0.799569 1.611478
OG_05637 3 1.192262 1.166134 1.165108 1.245543
OG_01452 3 1.191146 1.526632 0.922735 1.124072
OG_07153 3 1.174770 0.979025 1.358085 1.187201
OG_01414 3 1.161650 1.680689 0.775683 1.028577
OG_01753 3 1.160818 1.140792 1.097454 1.244207
OG_08663 3 1.138407 0.935220 1.247335 1.232668
OG_09796 3 1.133137 1.016265 1.074893 1.308253
OG_04467 3 1.121429 1.075908 1.254588 1.033790
OG_07892 3 1.113713 1.090661 0.877698 1.372780
============================================================
GENERATING VISUALIZATIONS
============================================================
/Users/sr320/GitHub/ConTra/code/run_biomin_species_comparison.py:397: FutureWarning:
Passing `palette` without assigning `hue` is deprecated and will be removed in v0.14.0. Assign the `x` variable to `hue` and set `legend=False` for the same effect.
sns.violinplot(data=strength_df, x='species', y='context_strength',
✅ Saved comparison plots: /Users/sr320/GitHub/ConTra/output/biomin_comparison_20251130_052506/plots/species_comparison_overview.png
✅ Saved comparison report: /Users/sr320/GitHub/ConTra/output/biomin_comparison_20251130_052506/cross_species_comparison_report.md
======================================================================
🎉 CROSS-SPECIES COMPARISON COMPLETE!
======================================================================
Results saved to: /Users/sr320/GitHub/ConTra/output/biomin_comparison_20251130_052506
======================================================================
```
## **What the script does:**
1. **Runs analysis on all 3 species** in [full-species-biomin](vscode-file://vscode-app/Applications/Visual%20Studio%20Code.app/Contents/Resources/app/out/vs/code/electron-browser/workbench/workbench.html):
- A. pulchra (apul)
- P. evermanni (peve)
- P. tuahiniensis (ptua)
2. **Extracts OG IDs** from the results using regex to parse formats like:
- `"('OG_13910', 'FUN_002435')"`
- `OG_13910`
3. **Compares results across species**:
- Finds OGs present in 2+ species
- Ranks by conservation and context strength
- Calculates correlations between species
4. **Generates outputs**:
- `cross_species_methylation_mirna_comparison.csv` - Full comparison table
- `plots/species_comparison_overview.png` - 4-panel visualization
- `cross_species_comparison_report.md` - Summary report
------------------------------------------------------------------------
# Cross-Species Biomin Context-Dependent Analysis Comparison
**Generated:** 2025-11-30 05:25:07
## Overview
This report compares context-dependent regulatory interactions across three coral species:
- **A. pulchra** (`apul`)
- **P. evermanni** (`peve`)
- **P. tuahiniensis** (`ptua`)
## Per-Species Summary
### A. pulchra (apul)
- Total methylation-miRNA interactions: 33270
- Context-dependent interactions: 3039
- Mean context strength: 0.311
- Unique OGs: 334
### P. evermanni (peve)
- Total methylation-miRNA interactions: 27150
- Context-dependent interactions: 3935
- Mean context strength: 0.371
- Unique OGs: 309
### P. tuahiniensis (ptua)
- Total methylation-miRNA interactions: 51150
- Context-dependent interactions: 8765
- Mean context strength: 0.346
- Unique OGs: 474
## Cross-Species Conservation
- **Total unique OGs:** 611
- **OGs in 2+ species:** 358 (58.6%)
- **OGs in all 3 species:** 100 (16.4%)
## Interpretation
OGs (orthologous groups) that show context-dependent regulation in multiple species suggest:
1. **Conserved regulatory mechanisms** - These genes may have fundamental roles where regulation is evolutionarily maintained
2. **Robust biological signals** - Conservation across species reduces the likelihood of false positives
3. **Candidates for functional validation** - These genes are high-priority targets for experimental follow-up
## Files Generated
- `cross_species_methylation_mirna_comparison.csv` - Full comparison table
- `plots/species_comparison_overview.png` - Visualization of cross-species patterns
- `cross_species_comparison_report.md` - This report
------------------------------------------------------------------------
```{r}
library(tidyverse)
library(reactable)
library(htmltools)
url <- "https://gannet.fish.washington.edu/v1_web/owlshell/bu-github/ConTra/output/biomin_comparison_20251130_052506/cross_species_methylation_mirna_comparison.csv"
df <- read_csv(url, show_col_types = FALSE)
# Format all numeric columns to 2 digits after decimal
df_fmt <- df %>%
mutate(across(where(is.numeric), ~ sprintf("%.2f", .x)))
reactable(
df_fmt,
searchable = TRUE,
filterable = TRUE,
pagination = TRUE,
highlight = TRUE,
striped = TRUE,
defaultPageSize = 20,
sortable = TRUE,
theme = reactable::reactableTheme(
highlightColor = "#EAF2F8"
)
)
```
# Ptua Context-Dependent Regulation Analysis Report
**Generated:** 2025-12-02 12:52:16
**Analysis ID:** 20251201_160901
## Executive Summary
This report presents the results of context-dependent regulatory interaction analysis between:
- Gene expression
- miRNA expression
- lncRNA expression
- DNA methylation
## Analysis Overview
- **Parallel Processing:** 192 CPU cores
- **Available RAM:** 2941.9 GB
- **Datasets Loaded:** 4 data types
- Gene: 32 samples × 483 features
- Lncrna: 32 samples × 11236 features
- Mirna: 32 samples × 40 features
- Methylation: 32 samples × 263324 features
## Results Summary
### Methylation-miRNA Context Analysis
- **Total interactions analyzed:** 33270
- **Context-dependent interactions (F-test):** 3039
- **Mean improvement from interaction:** 0.017
- **Mean context strength:** 0.311
#### Top 10 Methylation-miRNA Interactions (by Context Strength)
*This table shows genes whose regulation by DNA methylation is context-dependent on miRNA expression levels, ranked by context strength.*
| Rank | Gene | Methylation | miRNA | Improvement | Context Strength | Empirical FDR Sig. |
|------|------|-------------|-------|-------------|------------------|--------------------|
| 1 | ('OG_02537', 'Pocillopora_meandrina_HIv1___TS.g26115.t1') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000010_2525456 | mirna_Cluster_4609 | 0.043 | 1.805 | False |
| 2 | ('OG_08920', 'Pocillopora_meandrina_HIv1___RNAseq.g1090.t1') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000017_5066475 | mirna_Cluster_4094 | 0.016 | 1.770 | False |
| 3 | ('OG_01414', 'Pocillopora_meandrina_HIv1___TS.g28923.t1a') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000003_1369854 | mirna_Cluster_4826 | 0.015 | 1.681 | False |
| 4 | ('OG_04723', 'Pocillopora_meandrina_HIv1___RNAseq.g4434.t2') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000000_18894187 | mirna_Cluster_4826 | 0.131 | 1.600 | False |
| 5 | ('OG_15303', 'Pocillopora_meandrina_HIv1___TS.g5438.t1') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000018_8279715 | mirna_Cluster_6807 | 0.004 | 1.583 | False |
| 6 | ('OG_14804', 'Pocillopora_meandrina_HIv1___RNAseq.g24442.t1') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000006_12397256 | mirna_Cluster_4485 | 0.005 | 1.562 | False |
| 7 | ('OG_13910', 'Pocillopora_meandrina_HIv1___RNAseq.g778.t1') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000010_5065428 | mirna_Cluster_3415 | 0.002 | 1.561 | False |
| 8 | ('OG_08920', 'Pocillopora_meandrina_HIv1___RNAseq.g1090.t1') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000017_5066475 | mirna_Cluster_5123 | 0.040 | 1.545 | False |
| 9 | ('OG_01452', 'Pocillopora_meandrina_HIv1___RNAseq.g19573.t1') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000000_10732016 | mirna_Cluster_4826 | 0.265 | 1.527 | False |
| 10 | ('OG_01155', 'Pocillopora_meandrina_HIv1___RNAseq.g6698.t2') | methylation_CpG_Pocillopora_meandrina_HIv1___Sc0000011_3176997 | mirna_Cluster_4485 | 0.045 | 1.521 | False |
### lncRNA-miRNA Context Analysis (Exploratory)
- **Total interactions analyzed:** 33270
- **Context-dependent interactions (F-test):** 4724
- **Mean improvement from interaction:** 0.018
- **Mean context strength:** 0.229
#### Top 10 lncRNA-miRNA Interactions (Exploratory)
*This table identifies genes whose lncRNA-mediated regulation varies depending on miRNA expression. Note that current interaction metrics behave similarly or more strongly under randomized data and should be interpreted cautiously.*
| Rank | Gene | lncRNA | miRNA | Improvement | Context Strength |
|------|------|--------|-------|-------------|------------------|
| 1 | ('OG_14763', 'Pocillopora_meandrina_HIv1___RNAseq.g13186.t1') | lncrna_lncRNA_12271 | mirna_Cluster_6456 | 0.376 | nan |
| 2 | ('OG_14681', 'Pocillopora_meandrina_HIv1___TS.g17155.t1') | lncrna_lncRNA_30484 | mirna_Cluster_1159 | 0.360 | nan |
| 3 | ('OG_14763', 'Pocillopora_meandrina_HIv1___RNAseq.g13186.t1') | lncrna_lncRNA_12271 | mirna_Cluster_4 | 0.356 | nan |
| 4 | ('OG_14681', 'Pocillopora_meandrina_HIv1___TS.g17155.t1') | lncrna_lncRNA_5394 | mirna_Cluster_1159 | 0.354 | nan |
| 5 | ('OG_14681', 'Pocillopora_meandrina_HIv1___TS.g17155.t1') | lncrna_lncRNA_5394 | mirna_Cluster_3713 | 0.340 | nan |
| 6 | ('OG_14763', 'Pocillopora_meandrina_HIv1___RNAseq.g13186.t1') | lncrna_lncRNA_29138 | mirna_Cluster_6456 | 0.334 | nan |
| 7 | ('OG_14681', 'Pocillopora_meandrina_HIv1___TS.g17155.t1') | lncrna_lncRNA_30480 | mirna_Cluster_3713 | 0.333 | nan |
| 8 | ('OG_14763', 'Pocillopora_meandrina_HIv1___RNAseq.g13186.t1') | lncrna_lncRNA_29138 | mirna_Cluster_4 | 0.332 | nan |
| 9 | ('OG_14681', 'Pocillopora_meandrina_HIv1___TS.g17155.t1') | lncrna_lncRNA_30481 | mirna_Cluster_3713 | 0.329 | nan |
| 10 | ('OG_14763', 'Pocillopora_meandrina_HIv1___RNAseq.g13186.t1') | lncrna_lncRNA_29132 | mirna_Cluster_6456 | 0.326 | nan |
### Multi-Way Interaction Analysis (Exploratory)
- **Total genes analyzed:** 483
- **Genes with significant interactions (F-test):** 0
- **Mean improvement from interactions:** 0.597
#### Top 10 Multi-Way Interactions (Exploratory)
*This table shows genes with the largest multi-regulator improvements. Current evidence suggests similar improvements can arise under randomized data; treat these results as hypothesis-generating rather than definitive.*
| Rank | Gene | Improvement | Significant Interactions |
|------|------|-------------|------------------------|
| 1 | ('OG_07000', 'Pocillopora_meandrina_HIv1___RNAseq.g5328.t1') | 0.912 | False |
| 2 | ('OG_02317', 'Pocillopora_meandrina_HIv1___TS.g30016.t1') | 0.912 | False |
| 3 | ('OG_08545', 'Pocillopora_meandrina_HIv1___RNAseq.g1314.t1') | 0.912 | False |
| 4 | ('OG_00466', 'Pocillopora_meandrina_HIv1___TS.g4618.t1') | 0.911 | False |
| 5 | ('OG_07822', 'Pocillopora_meandrina_HIv1___RNAseq.g27709.t1') | 0.910 | False |
| 6 | ('OG_04209', 'Pocillopora_meandrina_HIv1___RNAseq.g10045.t1') | 0.908 | False |
| 7 | ('OG_01117', 'Pocillopora_meandrina_HIv1___TS.g20470.t1') | 0.907 | False |
| 8 | ('OG_17014', 'Pocillopora_meandrina_HIv1___RNAseq.g15484.t1') | 0.907 | False |
| 9 | ('OG_02666', 'Pocillopora_meandrina_HIv1___RNAseq.g21004.t1') | 0.906 | False |
| 10 | ('OG_07042', 'Pocillopora_meandrina_HIv1___TS.g12785.t1') | 0.906 | False |
## Data Files
<https://gannet.fish.washington.edu/v1_web/owlshell/bu-github/ConTra/output/context_dependent_analysis_20251201_160901/tables/>
The following data files were generated:
These CSV files contain the detailed results of context-dependent regulatory analysis, including interaction statistics, p-values, context strengths, and regulatory relationships for each analyzed gene-regulator pair.
- **high_methylation_gene_lncrna_correlations.csv** (130713.8 KB)
- **high_methylation_gene_methylation_correlations.csv** (1814558.8 KB)
- **high_methylation_gene_mirna_correlations.csv** (394.4 KB)
- **high_mirna_gene_lncrna_correlations.csv** (88440.6 KB)
- **high_mirna_gene_methylation_correlations.csv** (1713259.1 KB)
- **high_mirna_gene_mirna_correlations.csv** (377.4 KB)
- **lncrna_mirna_context.csv** (9879.4 KB)
- **low_mirna_gene_lncrna_correlations.csv** (187585.6 KB)
- **low_mirna_gene_methylation_correlations.csv** (1766258.4 KB)
- **low_mirna_gene_mirna_correlations.csv** (966.3 KB)
- **methylation_mirna_context.csv** (11173.1 KB)
- **multi_way_interactions.csv** (1218.8 KB)
### Table Definitions and Statistical Confidence
Below is a brief description of the main table types and how to assess statistical confidence for each:
- **methylation_mirna_context.csv**:
- One row per gene–CpG–miRNA triplet.
- Contains regression-based metrics (`r2_*`, `improvement_from_regulator2`, `improvement_from_interaction`), conditional correlations (`corr_high_regulator2`, `corr_low_regulator2`), and `context_strength`.
- When empirical FDR is enabled, includes `empirical_fdr_threshold`, `empirical_fdr_estimated`, and `empirical_fdr_significant`.
- **How to assess confidence**: prioritize interactions with `empirical_fdr_significant == True` (if available); otherwise, use `context_strength` as the primary effect-size metric, ideally cross-referenced against random-data null runs.
- **high_methylation_gene_methylation_correlations.csv** / **high_methylation_gene_mirna_correlations.csv** / **high_methylation_gene_lncrna_correlations.csv**:
- Correlation networks in the **high-methylation context**, defined using a sentinel CpG site.
- Each row is a gene–regulator pair with a Pearson correlation (`correlation`) and raw `p_value` computed only within high-methylation samples.
- **How to assess confidence**: these p-values are not corrected for multiple testing; treat them as exploratory. For higher confidence, focus on strong effect sizes (|correlation| close to 1) and/or overlap with FDR-supported context interactions from `methylation_mirna_context.csv`.
- **low_mirna_gene_methylation_correlations.csv** / **low_mirna_gene_mirna_correlations.csv** / **low_mirna_gene_lncrna_correlations.csv**:
- Correlation networks in the **low-miRNA context**, using a sentinel miRNA as the context-defining variable.
- Each row is a gene–regulator pair with a Pearson correlation (`correlation`) and raw `p_value` computed only within low-miRNA samples.
- **How to assess confidence**: as above, p-values are unadjusted across many tests; use them as a guide for ranking, not strict significance. Strong |correlation| and consistency with patterns seen in `methylation_mirna_context.csv` or across contexts provide more persuasive evidence.
- **lncrna_mirna_context.csv** (if lncRNA context module is enabled):
- Regression-based lncRNA–miRNA–gene context metrics analogous to `methylation_mirna_context.csv`.
- Empirical comparisons with randomized data suggest that current interaction metrics here are more exploratory; interpret "context-dependent" calls cautiously.
- **How to assess confidence**: use this table primarily for hypothesis generation or to find lncRNAs associated with genes that already have strong methylation–miRNA context signals.
- **multi_way_interactions.csv** (if multi-way module is enabled):
- Summarizes multi-regulator regression models for each gene, comparing a simple model to a full model with many regulators.
- Contains `improvement_from_regulators` and an F-test `interaction_p_value` with a boolean `has_significant_interactions`.
- **How to assess confidence**: current evidence shows that large improvements can also arise in randomized data; treat these results as exploratory and consider additional validation (e.g., overlap with simpler context metrics or external datasets).
## Analysis Parameters
- **Parallel workers:** 192
- **Data directory:** /mmfs1/gscratch/scrubbed/sr320/github/ConTra/data/full-species-biomin/cleaned_ptua
- **Output directory:** /mmfs1/gscratch/scrubbed/sr320/github/ConTra/output/context_dependent_analysis_20251201_160901
- **Analysis timestamp:** 20251201_160901
## Conclusion
This analysis successfully identified context-dependent regulatory interactions using optimized parallel processing.
The results provide insights into how different regulatory layers interact in a context-specific manner.
*Report generated by OptimizedContextDependentRegulationAnalysis on 2025-12-02 12:52:16*
# Peve Context-Dependent Regulation Analysis Report
**Generated:** 2025-11-30 19:41:04
**Analysis ID:** 20251130_091333
## Executive Summary
This report presents the results of context-dependent regulatory interaction analysis between:
- Gene expression
- miRNA expression
- lncRNA expression
- DNA methylation
## Analysis Overview
- **Parallel Processing:** 192 CPU cores
- **Available RAM:** 2930.5 GB
- **Datasets Loaded:** 4 data types
- Gene: 36 samples × 553 features
- Lncrna: 36 samples × 8319 features
- Mirna: 36 samples × 48 features
- Methylation: 36 samples × 104459 features
## Results Summary
### Methylation-miRNA Context Analysis
- **Total interactions analyzed:** 27150
- **Context-dependent interactions (F-test):** 3935
- **Mean improvement from interaction:** 0.022
- **Mean context strength:** 0.371
#### Top 10 Methylation-miRNA Interactions (by Context Strength)
*This table shows genes whose regulation by DNA methylation is context-dependent on miRNA expression levels, ranked by context strength.*
| Rank | Gene | Methylation | miRNA | Improvement | Context Strength | Empirical FDR Sig. |
|------|------|-------------|-------|-------------|------------------|--------------------|
| 1 | ('OG_11813', 'Peve_00021937') | methylation_CpG_Porites_evermani_scaffold_3933_26651 | mirna_Cluster_1004 | 0.028 | 1.523 | False |
| 2 | ('OG_11813', 'Peve_00021937') | methylation_CpG_Porites_evermani_scaffold_2342_55729 | mirna_Cluster_1004 | 0.043 | 1.513 | False |
| 3 | ('OG_11813', 'Peve_00021937') | methylation_CpG_Porites_evermani_scaffold_3933_26651 | mirna_Cluster_6236 | 0.052 | 1.455 | False |
| 4 | ('OG_08793', 'Peve_00007107') | methylation_CpG_Porites_evermani_scaffold_865_27253 | mirna_Cluster_3165 | 0.032 | 1.451 | False |
| 5 | ('OG_11813', 'Peve_00021937') | methylation_CpG_Porites_evermani_scaffold_3933_26651 | mirna_Cluster_4247 | 0.038 | 1.437 | False |
| 6 | ('OG_11360', 'Peve_00001484') | methylation_CpG_Porites_evermani_scaffold_325_68647 | mirna_Cluster_6234 | 0.130 | 1.388 | False |
| 7 | ('OG_06599', 'Peve_00038759') | methylation_CpG_Porites_evermani_scaffold_259_87918 | mirna_Cluster_9197 | 0.000 | 1.369 | False |
| 8 | ('OG_11360', 'Peve_00001484') | methylation_CpG_Porites_evermani_scaffold_325_68647 | mirna_Cluster_3165 | 0.164 | 1.362 | False |
| 9 | ('OG_11813', 'Peve_00021937') | methylation_CpG_Porites_evermani_scaffold_1152_147508 | mirna_Cluster_1004 | 0.054 | 1.360 | False |
| 10 | ('OG_07153', 'Peve_00000682') | methylation_CpG_Porites_evermani_scaffold_910_155276 | mirna_Cluster_5058 | 0.078 | 1.358 | False |
### lncRNA-miRNA Context Analysis (Exploratory)
- **Total interactions analyzed:** 27150
- **Context-dependent interactions (F-test):** 5550
- **Mean improvement from interaction:** 0.029
- **Mean context strength:** 0.282
#### Top 10 lncRNA-miRNA Interactions (Exploratory)
*This table identifies genes whose lncRNA-mediated regulation varies depending on miRNA expression. Note that current interaction metrics behave similarly or more strongly under randomized data and should be interpreted cautiously.*
| Rank | Gene | lncRNA | miRNA | Improvement | Context Strength |
|------|------|--------|-------|-------------|------------------|
| 1 | ('OG_18262', 'Peve_00044240') | lncrna_lncRNA_10659 | mirna_Cluster_6099 | 0.469 | 1.007 |
| 2 | ('OG_18262', 'Peve_00044240') | lncrna_lncRNA_10659 | mirna_Cluster_6918 | 0.458 | 0.989 |
| 3 | ('OG_13684', 'Peve_00018981') | lncrna_lncRNA_12581 | mirna_Cluster_6234 | 0.444 | 0.898 |
| 4 | ('OG_12320', 'Peve_00024570') | lncrna_lncRNA_19767 | mirna_Cluster_670 | 0.440 | nan |
| 5 | ('OG_12320', 'Peve_00024570') | lncrna_lncRNA_19765 | mirna_Cluster_670 | 0.439 | nan |
| 6 | ('OG_13526', 'Peve_00019641') | lncrna_lncRNA_20807 | mirna_Cluster_670 | 0.437 | nan |
| 7 | ('OG_13526', 'Peve_00019641') | lncrna_lncRNA_23227 | mirna_Cluster_670 | 0.437 | nan |
| 8 | ('OG_13684', 'Peve_00018981') | lncrna_lncRNA_23131 | mirna_Cluster_981 | 0.433 | 0.861 |
| 9 | ('OG_13684', 'Peve_00018981') | lncrna_lncRNA_12581 | mirna_Cluster_5257 | 0.433 | 0.975 |
| 10 | ('OG_13684', 'Peve_00018981') | lncrna_lncRNA_12581 | mirna_Cluster_6235 | 0.432 | 0.987 |
### Multi-Way Interaction Analysis (Exploratory)
- **Total genes analyzed:** 553
- **Genes with significant interactions (F-test):** 0
- **Mean improvement from interactions:** 0.574
#### Top 10 Multi-Way Interactions (Exploratory)
*This table shows genes with the largest multi-regulator improvements. Current evidence suggests similar improvements can arise under randomized data; treat these results as hypothesis-generating rather than definitive.*
| Rank | Gene | Improvement | Significant Interactions |
|------|------|-------------|------------------------|
| 1 | ('OG_13066', 'Peve_00003034') | 0.921 | False |
| 2 | ('OG_13685', 'Peve_00034973') | 0.921 | False |
| 3 | ('OG_10907', 'Peve_00033654') | 0.921 | False |
| 4 | ('OG_05950', 'Peve_00044664') | 0.921 | False |
| 5 | ('OG_01981', 'Peve_00033650') | 0.921 | False |
| 6 | ('OG_03267', 'Peve_00013805') | 0.920 | False |
| 7 | ('OG_17015', 'Peve_00023327') | 0.919 | False |
| 8 | ('OG_02088', 'Peve_00026747') | 0.919 | False |
| 9 | ('OG_05713', 'Peve_00041095') | 0.919 | False |
| 10 | ('OG_12805', 'Peve_00005899') | 0.919 | False |
## Data Files
<https://gannet.fish.washington.edu/v1_web/owlshell/bu-github/ConTra/output/context_dependent_analysis_20251130_091333/tables/>
The following data files were generated:
These CSV files contain the detailed results of context-dependent regulatory analysis, including interaction statistics, p-values, context strengths, and regulatory relationships for each analyzed gene-regulator pair.
- **high_methylation_gene_lncrna_correlations.csv** (69461.9 KB)
- **high_methylation_gene_methylation_correlations.csv** (675831.9 KB)
- **high_methylation_gene_mirna_correlations.csv** (262.4 KB)
- **high_mirna_gene_lncrna_correlations.csv** (65272.2 KB)
- **high_mirna_gene_methylation_correlations.csv** (604621.8 KB)
- **high_mirna_gene_mirna_correlations.csv** (144.7 KB)
- **lncrna_mirna_context.csv** (6740.5 KB)
- **low_mirna_gene_lncrna_correlations.csv** (111381.4 KB)
- **low_mirna_gene_methylation_correlations.csv** (641498.3 KB)
- **low_mirna_gene_mirna_correlations.csv** (171.5 KB)
- **methylation_mirna_context.csv** (7651.1 KB)
- **multi_way_interactions.csv** (1213.2 KB)
### Table Definitions and Statistical Confidence
Below is a brief description of the main table types and how to assess statistical confidence for each:
- **methylation_mirna_context.csv**:
- One row per gene–CpG–miRNA triplet.
- Contains regression-based metrics (`r2_*`, `improvement_from_regulator2`, `improvement_from_interaction`), conditional correlations (`corr_high_regulator2`, `corr_low_regulator2`), and `context_strength`.
- When empirical FDR is enabled, includes `empirical_fdr_threshold`, `empirical_fdr_estimated`, and `empirical_fdr_significant`.
- **How to assess confidence**: prioritize interactions with `empirical_fdr_significant == True` (if available); otherwise, use `context_strength` as the primary effect-size metric, ideally cross-referenced against random-data null runs.
- **high_methylation_gene_methylation_correlations.csv** / **high_methylation_gene_mirna_correlations.csv** / **high_methylation_gene_lncrna_correlations.csv**:
- Correlation networks in the **high-methylation context**, defined using a sentinel CpG site.
- Each row is a gene–regulator pair with a Pearson correlation (`correlation`) and raw `p_value` computed only within high-methylation samples.
- **How to assess confidence**: these p-values are not corrected for multiple testing; treat them as exploratory. For higher confidence, focus on strong effect sizes (|correlation| close to 1) and/or overlap with FDR-supported context interactions from `methylation_mirna_context.csv`.
- **low_mirna_gene_methylation_correlations.csv** / **low_mirna_gene_mirna_correlations.csv** / **low_mirna_gene_lncrna_correlations.csv**:
- Correlation networks in the **low-miRNA context**, using a sentinel miRNA as the context-defining variable.
- Each row is a gene–regulator pair with a Pearson correlation (`correlation`) and raw `p_value` computed only within low-miRNA samples.
- **How to assess confidence**: as above, p-values are unadjusted across many tests; use them as a guide for ranking, not strict significance. Strong |correlation| and consistency with patterns seen in `methylation_mirna_context.csv` or across contexts provide more persuasive evidence.
- **lncrna_mirna_context.csv** (if lncRNA context module is enabled):
- Regression-based lncRNA–miRNA–gene context metrics analogous to `methylation_mirna_context.csv`.
- Empirical comparisons with randomized data suggest that current interaction metrics here are more exploratory; interpret "context-dependent" calls cautiously.
- **How to assess confidence**: use this table primarily for hypothesis generation or to find lncRNAs associated with genes that already have strong methylation–miRNA context signals.
- **multi_way_interactions.csv** (if multi-way module is enabled):
- Summarizes multi-regulator regression models for each gene, comparing a simple model to a full model with many regulators.
- Contains `improvement_from_regulators` and an F-test `interaction_p_value` with a boolean `has_significant_interactions`.
- **How to assess confidence**: current evidence shows that large improvements can also arise in randomized data; treat these results as exploratory and consider additional validation (e.g., overlap with simpler context metrics or external datasets).
## Analysis Parameters
- **Parallel workers:** 192
- **Data directory:** /mmfs1/gscratch/scrubbed/sr320/github/ConTra/data/full-species-biomin/cleaned_peve
- **Output directory:** /mmfs1/gscratch/scrubbed/sr320/github/ConTra/output/context_dependent_analysis_20251130_091333
- **Analysis timestamp:** 20251130_091333
## Conclusion
This analysis successfully identified context-dependent regulatory interactions using optimized parallel processing.
The results provide insights into how different regulatory layers interact in a context-specific manner.
*Report generated by OptimizedContextDependentRegulationAnalysis on 2025-11-30 19:41:04*
# Apul Context-Dependent Regulation Analysis Report
**Generated:** 2025-12-01 11:13:07
**Analysis ID:** 20251201_061836
## Executive Summary
This report presents the results of context-dependent regulatory interaction analysis between:
- Gene expression
- miRNA expression
- lncRNA expression
- DNA methylation
## Analysis Overview
- **Parallel Processing:** 192 CPU cores
- **Available RAM:** 2940.2 GB
- **Datasets Loaded:** 4 data types
- Gene: 39 samples × 517 features
- Lncrna: 39 samples × 15559 features
- Mirna: 39 samples × 51 features
- Methylation: 39 samples × 66428 features
## Results Summary
### Methylation-miRNA Context Analysis
- **Total interactions analyzed:** 51150
- **Context-dependent interactions (F-test):** 8765
- **Mean improvement from interaction:** 0.027
- **Mean context strength:** 0.346
#### Top 10 Methylation-miRNA Interactions (by Context Strength)
*This table shows genes whose regulation by DNA methylation is context-dependent on miRNA expression levels, ranked by context strength.*
| Rank | Gene | Methylation | miRNA | Improvement | Context Strength | Empirical FDR Sig. |
|------|------|-------------|-------|-------------|------------------|--------------------|
| 1 | ('OG_05303', 'FUN_023974') | methylation_CpG_ptg000001l_9710125 | mirna_Cluster_3109 | 0.060 | 1.881 | False |
| 2 | ('OG_07265', 'FUN_033056') | methylation_CpG_ptg000021l_10139814 | mirna_Cluster_3109 | 0.018 | 1.875 | False |
| 3 | ('OG_14547', 'FUN_019026') | methylation_CpG_ptg000023l_11736283 | mirna_Cluster_3109 | 0.023 | 1.875 | False |
| 4 | ('OG_07265', 'FUN_033056') | methylation_CpG_ptg000018l_7377398 | mirna_Cluster_3109 | 0.067 | 1.854 | False |
| 5 | ('OG_10509', 'FUN_002531') | methylation_CpG_ptg000031l_5310815 | mirna_Cluster_3109 | 0.004 | 1.837 | False |
| 6 | ('OG_06610', 'FUN_029760') | methylation_CpG_ptg000031l_5310762 | mirna_Cluster_3109 | 0.004 | 1.833 | False |
| 7 | ('OG_10811', 'FUN_006888') | methylation_CpG_ptg000023l_8252944 | mirna_Cluster_3109 | 0.006 | 1.756 | False |
| 8 | ('OG_14711', 'FUN_024326') | methylation_CpG_ptg000012l_17885415 | mirna_Cluster_3109 | 0.045 | 1.672 | False |
| 9 | ('OG_10811', 'FUN_006888') | methylation_CpG_ptg000027l_7575122 | mirna_Cluster_3109 | 0.015 | 1.660 | False |
| 10 | ('OG_06610', 'FUN_029760') | methylation_CpG_ptg000021l_6263529 | mirna_Cluster_3109 | 0.012 | 1.657 | False |
### lncRNA-miRNA Context Analysis (Exploratory)
- **Total interactions analyzed:** 51150
- **Context-dependent interactions (F-test):** 8365
- **Mean improvement from interaction:** 0.021
- **Mean context strength:** 0.270
#### Top 10 lncRNA-miRNA Interactions (Exploratory)
*This table identifies genes whose lncRNA-mediated regulation varies depending on miRNA expression. Note that current interaction metrics behave similarly or more strongly under randomized data and should be interpreted cautiously.*
| Rank | Gene | lncRNA | miRNA | Improvement | Context Strength |
|------|------|--------|-------|-------------|------------------|
| 1 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_48837 | mirna_Cluster_14165 | 0.420 | 1.162 |
| 2 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_48837 | mirna_Cluster_14146 | 0.407 | 1.082 |
| 3 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_48837 | mirna_Cluster_3226 | 0.406 | 0.969 |
| 4 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_48837 | mirna_Cluster_10452 | 0.400 | 0.910 |
| 5 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_23415 | mirna_Cluster_9512 | 0.394 | 1.033 |
| 6 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_23414 | mirna_Cluster_9512 | 0.392 | 1.021 |
| 7 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_23416 | mirna_Cluster_9512 | 0.391 | 1.028 |
| 8 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_54028 | mirna_Cluster_4752 | 0.387 | 0.966 |
| 9 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_48837 | mirna_Cluster_4752 | 0.379 | 1.288 |
| 10 | ('OG_10579', 'FUN_003891') | lncrna_lncRNA_44695 | mirna_Cluster_4752 | 0.377 | 0.834 |
### Multi-Way Interaction Analysis (Exploratory)
- **Total genes analyzed:** 517
- **Genes with significant interactions (F-test):** 0
- **Mean improvement from interactions:** 0.750
#### Top 10 Multi-Way Interactions (Exploratory)
*This table shows genes with the largest multi-regulator improvements. Current evidence suggests similar improvements can arise under randomized data; treat these results as hypothesis-generating rather than definitive.*
| Rank | Gene | Improvement | Significant Interactions |
|------|------|-------------|------------------------|
| 1 | ('OG_10450', 'FUN_001498') | 0.928 | False |
| 2 | ('OG_12424', 'FUN_030063') | 0.928 | False |
| 3 | ('OG_10523', 'FUN_002672') | 0.928 | False |
| 4 | ('OG_12286', 'FUN_028190') | 0.926 | False |
| 5 | ('OG_09897', 'FUN_042992') | 0.925 | False |
| 6 | ('OG_14500', 'FUN_017790') | 0.925 | False |
| 7 | ('OG_04209', 'FUN_016791') | 0.923 | False |
| 8 | ('OG_10759', 'FUN_006140') | 0.922 | False |
| 9 | ('OG_08920', 'FUN_038376') | 0.921 | False |
| 10 | ('OG_02176', 'FUN_008525') | 0.920 | False |
## Data Files
<https://gannet.fish.washington.edu/v1_web/owlshell/bu-github/ConTra/output/context_dependent_analysis_20251201_061836/tables/>
The following data files were generated:
These CSV files contain the detailed results of context-dependent regulatory analysis, including interaction statistics, p-values, context strengths, and regulatory relationships for each analyzed gene-regulator pair.
- **high_methylation_gene_lncrna_correlations.csv** (130703.2 KB)
- **high_methylation_gene_methylation_correlations.csv** (284276.8 KB)
- **high_methylation_gene_mirna_correlations.csv** (177.7 KB)
- **lncrna_mirna_context.csv** (13579.9 KB)
- **low_mirna_gene_lncrna_correlations.csv** (96572.2 KB)
- **low_mirna_gene_methylation_correlations.csv** (343631.8 KB)
- **low_mirna_gene_mirna_correlations.csv** (266.4 KB)
- **methylation_mirna_context.csv** (14423.3 KB)
- **multi_way_interactions.csv** (968.1 KB)
### Table Definitions and Statistical Confidence
Below is a brief description of the main table types and how to assess statistical confidence for each:
- **methylation_mirna_context.csv**:
- One row per gene–CpG–miRNA triplet.
- Contains regression-based metrics (`r2_*`, `improvement_from_regulator2`, `improvement_from_interaction`), conditional correlations (`corr_high_regulator2`, `corr_low_regulator2`), and `context_strength`.
- When empirical FDR is enabled, includes `empirical_fdr_threshold`, `empirical_fdr_estimated`, and `empirical_fdr_significant`.
- **How to assess confidence**: prioritize interactions with `empirical_fdr_significant == True` (if available); otherwise, use `context_strength` as the primary effect-size metric, ideally cross-referenced against random-data null runs.
- **high_methylation_gene_methylation_correlations.csv** / **high_methylation_gene_mirna_correlations.csv** / **high_methylation_gene_lncrna_correlations.csv**:
- Correlation networks in the **high-methylation context**, defined using a sentinel CpG site.
- Each row is a gene–regulator pair with a Pearson correlation (`correlation`) and raw `p_value` computed only within high-methylation samples.
- **How to assess confidence**: these p-values are not corrected for multiple testing; treat them as exploratory. For higher confidence, focus on strong effect sizes (|correlation| close to 1) and/or overlap with FDR-supported context interactions from `methylation_mirna_context.csv`.
- **low_mirna_gene_methylation_correlations.csv** / **low_mirna_gene_mirna_correlations.csv** / **low_mirna_gene_lncrna_correlations.csv**:
- Correlation networks in the **low-miRNA context**, using a sentinel miRNA as the context-defining variable.
- Each row is a gene–regulator pair with a Pearson correlation (`correlation`) and raw `p_value` computed only within low-miRNA samples.
- **How to assess confidence**: as above, p-values are unadjusted across many tests; use them as a guide for ranking, not strict significance. Strong |correlation| and consistency with patterns seen in `methylation_mirna_context.csv` or across contexts provide more persuasive evidence.
- **lncrna_mirna_context.csv** (if lncRNA context module is enabled):
- Regression-based lncRNA–miRNA–gene context metrics analogous to `methylation_mirna_context.csv`.
- Empirical comparisons with randomized data suggest that current interaction metrics here are more exploratory; interpret "context-dependent" calls cautiously.
- **How to assess confidence**: use this table primarily for hypothesis generation or to find lncRNAs associated with genes that already have strong methylation–miRNA context signals.
- **multi_way_interactions.csv** (if multi-way module is enabled):
- Summarizes multi-regulator regression models for each gene, comparing a simple model to a full model with many regulators.
- Contains `improvement_from_regulators` and an F-test `interaction_p_value` with a boolean `has_significant_interactions`.
- **How to assess confidence**: current evidence shows that large improvements can also arise in randomized data; treat these results as exploratory and consider additional validation (e.g., overlap with simpler context metrics or external datasets).
## Analysis Parameters
- **Parallel workers:** 192
- **Data directory:** /mmfs1/gscratch/scrubbed/sr320/github/ConTra/data/full-species-biomin/cleaned_apul
- **Output directory:** /mmfs1/gscratch/scrubbed/sr320/github/ConTra/output/context_dependent_analysis_20251201_061836
- **Analysis timestamp:** 20251201_061836
## Conclusion
This analysis successfully identified context-dependent regulatory interactions using optimized parallel processing.
The results provide insights into how different regulatory layers interact in a context-specific manner.
*Report generated by OptimizedContextDependentRegulationAnalysis on 2025-12-01 11:13:07*
