Grassmannian Geodesic Distance Predicts Cross-Cohort Classifier Degradation Under Analytical Heterogeneity, After Controlling for Source Classifier Quality | Zenodo

Skip to main

You are using an outdated browser. Please upgrade your browser to improve your experience.

Published July 7, 2026

| Version v1

Preprint

Open

Grassmannian Geodesic Distance Predicts Cross-Cohort Classifier Degradation Under Analytical Heterogeneity, After Controlling for Source Classifier Quality

Authors/Creators

Tower, Elliot

Description

Geodesic distance on the Grassmannian manifold Gr(k, d) between cohort-specific PCA subspaces, after partialing out source classifier quality, predicts the residual AUC gap when a classifier is transferred to a new cohort. Raw geodesic–gap correlations are null on all seven datasets because the AUC gap conflates source internal performance with distribution shift. The confound-control step is essential.

The method is validated on a colorectal cancer microbiome meta-analysis (9 studies, 824 samples; partial rho = +0.61, clustered bootstrap 95% CI [−0.02, +0.80], 97% positive; Delta R^2 = +0.23) and on the QMDiab multi-biofluid metabolomics study (3 biofluids × 3 ethnicities, 356 participants; partial rho = +0.40, clustered CI [+0.04, +0.73]). Leave-one-study-out validation confirms out-of-sample utility (42% MAE reduction over baseline). QMDiab decomposition shows the signal reflects the contrast between cross-biofluid pairs (partial rho = −0.04) and within-biofluid cross-ethnicity pairs (partial rho = +0.14), consistent with the boundary-condition thesis rather than graded within-type prediction.

Five additional datasets (IBD microbiome, SPIROMICS COPD, breast cancer GEO, TCGA-BRCA, MTBLS7260 metabolomics) yield null results, identifying interpretable boundary conditions: the method requires analytical heterogeneity between cohorts. Centralized platforms or shared microarray platforms produce null results even across independent study sites. A formal power analysis shows 41% power to detect the CRC effect at the effective sample size of 9 studies; 19 studies would be needed for 80% power.

This deposit contains the PLOS ONE submission manuscript (LaTeX + PDF), cover letter, pre-registration document (frozen at commit 675290c before confirmatory analyses), confirmatory analysis scripts (power analysis, QMDiab decomposition), the full analysis codebase (Python), and all result files (JSON summaries, ordered-pair CSVs, figures) for all seven datasets. Raw data are publicly available from their respective repositories (curatedMetagenomicData, MetaboLights MTBLS59, GEO, TCGA) and are not included due to size.

Files

grassmannian_cohort_transfer.pdf

Files<br>(4.4 MB)

Name<br>Size

Download all

grassmannian_cohort_transfer.pdf

md5:2c5ebfe94237e2c05a087c549da7d4df

1.0 MB

Preview

Download

grassmannian_cohort_transfer.zip

md5:18861be0ca9d4e40860805bacf2ee346

3.3 MB

Preview

Download

Additional details

Software

Repository URL

https://github.com/elliottower/biomedical-cohort-transfer

Views

Downloads

Show more details

All versions<br>This version

Views

Total views

Downloads

Total downloads

Data volume

Total data volume

2.0 MB<br>2.0 MB

More info on how stats are collected....

Versions

External resources

Indexed in

OpenAIRE

Communities

Keywords and subjects

Keywords

Grassmannian manifold

cross-cohort transfer

classifier degradation

PCA subspaces

geodesic distance

partial correlation

microbiome

metabolomics

analytics heterogeneity

distribution shift

MeSH

Computational Biology

Machine Learning

Colorectal Neoplasms

Metabolomics

Principal Component Analysis

Models, Statistical

Reproducibility of Results

Cohort Studies

Details

DOI

DOI Badge

DOI

10.5281/zenodo.21248764

Markdown

[](https://doi.org/10.5281/zenodo.21248764)

reStructuredText

.. image:: https://zenodo.org/badge/DOI/10.5281/zenodo.21248764.svg<br>:target: https://doi.org/10.5281/zenodo.21248764

HTML

Image URL

https://zenodo.org/badge/DOI/10.5281/zenodo.21248764.svg

Target URL

https://doi.org/10.5281/zenodo.21248764

Resource type<br>Preprint

Publisher<br>Zenodo

Rights

License

Creative Commons Attribution 4.0 International

The Creative Commons Attribution license allows re-distribution and re-use of a licensed work on the condition that the creator is appropriately credited.

Read more

Citation

Export

Technical metadata

Created

July 7, 2026

Modified

July 7, 2026

Jump up

This site uses cookies. Find out more on how we use cookies

Accept all cookies<br>Accept only essential cookies