---
id: "claim:normalization-for-sampled-count-data-2026-06-07-biorxiv-v2:658918466ec94021068e"
kind: "claim"
title: "Feature-panel stability used top-750 and top-1000 genes and Procrustes-aligned ten-PC embeddings."
description: ""
url: "https://sina.bio/claim/feature-panel-stability-used-top-750-and-top-1000-genes-and-procrustes-88f0bfe9"
---

# Feature-panel stability used top-750 and top-1000 genes and Procrustes-aligned ten-PC embeddings.

## Content

Feature-panel stability used top-750 and top-1000 genes and Procrustes-aligned ten-PC embeddings.

## Metadata

```json
{
  "claim_position": 54,
  "source_publication_id": "version:normalization-for-sampled-count-data--2026-06-07-biorxiv-v2",
  "source_work_id": "work:normalization-for-sampled-count-data"
}
```

## Connections (1)

### Depth normalization for single-cell genomics count data

- ID: `version:normalization-for-sampled-count-data--2026-06-07-biorxiv-v2`
- Kind: `publication`
- Direction: `incoming`
- URL: <https://sina.bio/publication/depth-normalization-for-single-cell-genomics-count-data-dcc8cf45>

This publication contains this grounded claim.

> We then formed the top 750 and top 1000 gene panels.

Source: 2026-06-07-biorxiv-v2.pdf — <https://sina.bio/source/2026-06-07-biorxiv-v2-pdf>

> The 1000-gene PCA
> embedding was centered, scaled, and orthogonally Procrustes-
> aligned to the 750-gene embedding in the top 10 PCs.

Source: 2026-06-07-biorxiv-v2.pdf — <https://sina.bio/source/2026-06-07-biorxiv-v2-pdf>
