Skip to content

Add VEN spaceflight cognition analysis code and figures #1

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
esila-keskin wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Add VEN spaceflight cognition analysis code and figures #1

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
48 changes: 48 additions & 0 deletions VEN_Spaceflight_Cognition/.gitignore
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,48 @@
# Large raw data files (not tracked -- obtain from NCBI GEO / NASA OSDR)
GSE259421_all_counts.txt
data/raw/GLDS-202_rna_seq_differential_expression_GLbulkRNAseq.csv

# Proteomics raw data (obtain from PRIDE PXD069807)
*.sepr2
ISS_WT83_proteins.txt
Ground_WT83_proteins.txt

# PDF output files (regenerate from source)
figures/*.pdf
*.pdf

# Results images from exploratory analysis (not paper figures)
results/*.png

# Python cache
__pycache__/
*.pyc
*.pyo
.ipynb_checkpoints/

# LaTeX build artefacts
*.aux
*.log
*.bbl
*.blg
*.out
*.fls
*.fdb_latexmk
*.synctex.gz
*.toc
*.nav
*.snm

# Office temp files
~$*

# OS files
.DS_Store
Thumbs.db
desktop.ini

# Editor files
*.swp
*.swo
.vscode/
.idea/
158 changes: 158 additions & 0 deletions VEN_Spaceflight_Cognition/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,158 @@
# VEN Fatigue Hypothesis: Social Cognition Paradox in Long-Duration Spaceflight

**Paper:** "The Social Cognition Paradox in Long-Duration Spaceflight: A VEN Fatigue Hypothesis for Duration-Dependent Emotion Recognition Decline"
**Authors:** Esila Keskin (UWE Bristol), Nilufar Ali (Boise State / NASA GeneLab Brain AWG), Margaret Windy McNerney (Stanford)

---

## Overview

This repository contains all analysis code and data (where distributable) for the paper.

The study derives and tests the **VEN Fatigue Hypothesis**: Von Economo neurons (VENs), optimised for fast social decisions on minimal evidence, are overactivated by gravity-disrupted social cues during spaceflight. This drives a compensatory myelination upregulation that sustains Emotion Recognition Task (ERT) speed within 6-month missions. Beyond this threshold, fatigue accumulates faster than compensation, producing selective late-inflight ERT speed decline.

Seven convergent lines of evidence are tested across five independent datasets spanning cognitive, transcriptomic, proteomic, and human blood-level analyses.

---

## Seven lines of evidence

1. **ERT cognitive paradox** -- ERT speed is stable across 6-month ISS missions (N=24; +0.106 SD) but shows the largest early-to-late inflight decline of any cognitive speed domain in the 340-day NASA Twins Study (N=1; -1.8 SD; within-individual cross-domain t=11.41, df=8)
2. **ISS rodent myelination** -- Myelination genes are specifically upregulated in ISS frontal cortex (GSE239336; +0.381 log2FC; 3.89 SD above genome-wide null; permutation p=0.0001)
3. **Ground analogue control** -- No specific myelination signal in OSD-202 hindlimb unloading + radiation (permutation p=0.164), with directional reversal consistent with a qualitatively distinct molecular response
4. **Human cortical organoids** -- Layer V projection genes massively upregulated in ISS cortical organoids (GSE259421; +1.347 log2FC; 6.17 SD above null; permutation p=0.0008)
5. **Dopaminergic organoid replication** -- Layer V signal replicates in dopaminergic (midbrain) organoids from the same mission (7.12 SD; p=0.0002), confirming the response is not cortical-lineage-specific
6. **Proteomics (PXD069807)** -- VEN-associated structural proteins in iPSC organoid proteomics (Jourdon et al. 2026) follow global ISS trends; Layer V transcription factors fall below mass-spectrometry detection threshold as predicted
7. **SOMA atlas human blood** -- All eight queried VEN panel genes reach p<0.05 in human astronaut blood (NASA Twins Study + Inspiration4); MOG log2FC +5.82 (p=4.59e-27) in year-long ISS, replicating in Inspiration4 PBMC (MBP log2FC +0.60)

---

## Repository structure

```
ven-spaceflight-cognition/

figures/ -- Publication figures (PNG)
fig1_ert_paradox.png
fig2_molecular_dissociation.png
fig3_organoid_permutation.png
fig4_domain_specificity.png
fig5_soma_blood.png
fig5_proteomics_permutation.png
dopaminergic_vs_cortical_comparison.png

data/raw/
twins_cognitive_heatmap.csv -- Digitised from Garrett-Bakelman et al. 2019, Fig 10B
dev2024_raw_scores.csv -- From Dev et al. 2024, Table 2
GSE239336_FCT_GCvsFLT-SAL_DEanalysis.txt -- ISS rodent DE analysis (NCBI GEO)
GLDS-202_...csv -- OSD-202 (NASA OSDR) [gitignored, 134 MB]
GSE259421_all_counts.txt -- Human organoid counts (NCBI GEO) [gitignored, 45 MB]

results/ -- All analysis outputs (CSV, JSON)

step1_create_cognitive_csvs.py -- Creates cognitive data CSVs from digitised values
step2_run_analysis.py -- ISS rodent (GSE239336) and OSD-202 analyses
step2_v2_run_analysis.py -- Updated molecular analysis pipeline
step3_make_figures.py -- Generates publication figures
make_paper_figures.py -- Updated figure generation script
ven_organoid_analysis.py -- Core cortical organoid VEN panel pipeline
ven_dopaminergic_analysis.py -- Dopaminergic organoid replication analysis
soma_analysis.py -- SOMA atlas VEN panel survey (human astronaut blood)
alysson_proteomics_analysis.py -- PatternLab V export parser for PXD069807 proteomics
parse_sepr2.py -- MS-NRBF binary parser for .sepr2 PatternLab files
inspect_sepr2.py -- Diagnostic utility for .sepr2 binary structure
```

---

## Data availability

| Dataset | Source | Size | Status |
|---------|--------|------|--------|
| `twins_cognitive_heatmap.csv` | Digitised from Garrett-Bakelman et al. 2019, Science | 1 KB | Included |
| `dev2024_raw_scores.csv` | Dev et al. 2024, Table 2 | 1 KB | Included |
| `GSE239336_FCT_GCvsFLT-SAL_DEanalysis.txt` | NCBI GEO: GSE239336 | ~1 MB | Included |
| `GLDS-202_rna_seq_...csv` | NASA OSDR: OSD-202 | 134 MB | Gitignored -- download from [NASA OSDR](https://osdr.nasa.gov) |
| `GSE259421_all_counts.txt` | NCBI GEO: GSE259421 | 45 MB | Gitignored -- download from [NCBI GEO](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE259421) |
| `ISS_WT83_proteins.txt` / `Ground_WT83_proteins.txt` | PRIDE: PXD069807 (Jourdon et al. 2026) | varies | Gitignored -- download from [PRIDE](https://www.ebi.ac.uk/pride/archive/projects/PXD069807) |
| SOMA atlas queries | soma.weill.cornell.edu (Overbey et al. Nature 2024) | N/A | Values extracted manually; hardcoded in `soma_analysis.py` |

---

## Reproducing the analysis

### 1. Install dependencies

```bash
pip install numpy pandas scipy matplotlib mygene
```

### 2. Download large data files (gitignored)

- OSD-202: download `GLDS-202_rna_seq_differential_expression_GLbulkRNAseq.csv` from [NASA OSDR](https://osdr.nasa.gov), place in `data/raw/`
- GSE259421: download `GSE259421_all_counts.txt` from [NCBI GEO](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE259421), place in the root directory
- PXD069807: download PatternLab V exports (`ISS_WT83_proteins.txt`, `Ground_WT83_proteins.txt`) from [PRIDE PXD069807](https://www.ebi.ac.uk/pride/archive/projects/PXD069807), place in the root directory

### 3. Run the analysis pipeline

```bash
# Step 1: create cognitive performance CSVs from digitised values
python step1_create_cognitive_csvs.py

# Step 2a: ISS rodent transcriptomics (GSE239336) and OSD-202 ground analogue
python step2_run_analysis.py

# Step 2b: human cortical organoid VEN panel analysis (GSE259421)
python ven_organoid_analysis.py

# Step 2c: dopaminergic organoid replication
python ven_dopaminergic_analysis.py

# Step 2d: proteomics analysis (PXD069807)
python alysson_proteomics_analysis.py

# Step 2e: SOMA atlas human blood survey
python soma_analysis.py

# Step 3: generate publication figures
python make_paper_figures.py
```

Results are saved to `results/` and figures to `figures/`.

---

## VEN gene panel

The 31-gene panel was defined in [Keskin 2026 (arXiv:2604.09229)](https://arxiv.org/abs/2604.09229) prior to accessing any dataset analysed here. All directional predictions were pre-registered via the NASA OSDR Brain AWG forum (April 2026).

| Category | Genes |
|----------|-------|
| Myelination (n=7) | MBP, MOG, PLP1, MAG, CNP, MOBP, ERMN |
| Fast Signalling (n=7) | SCN1A, KCNQ2, ANK3, NEFH, NEFM, NEFL, SNCG |
| Social Circuit (n=6) | OXTR, AVPR1A, HTR2A, DRD1, CHRM1, GABRB2 |
| Layer V Projection (n=5) | FEZF2, BCL11B, TBR1, SATB2, CUX1 |
| Metabolic Support (n=6) | VDAC1, ATP2B2, SLC17A7, SNAP25, SYP, NRXN1 |

NOS1 (direct VEN biochemical marker) tracked individually outside the panel.

Statistical approach: one-sample t-tests against zero and permutation tests (N=10,000 random gene sets of equal size drawn from the full expressed transcriptome; seed 42). Results reported as standard deviations above the genome-wide permutation null alongside two-tailed permutation p-values.

---

## Citation

If you use this code or data, please cite:

> Keskin E, Ali N, McNerney MW. The Social Cognition Paradox in Long-Duration Spaceflight: A VEN Fatigue Hypothesis for Duration-Dependent Emotion Recognition Decline. 2026. Preprint.

Related preprint (Fast Lane Hypothesis SNN model):

> Keskin E. Fast Lane Hypothesis: Von Economo Neurons as a Biological Speed-Accuracy Tradeoff Mechanism. arXiv:2604.09229. 2026.

---

## License

Code: MIT License. See LICENSE file.
Data: subject to original dataset terms (NCBI GEO / NASA OSDR open data policies; PRIDE proteomics repository terms).
Loading