﻿Available data and code for the second manuscript revision

The archive contains source records supplied by the authors, with updates from
the recovered single-cell RDS. The original project scripts and data are retained
unchanged. They are archival evidence, not a claim of a successful clean rerun.

New_Verification/ contains:
- cell_metadata.csv and saved object command history from the recovered RDS;
- RNA layer statistics and count-library totals;
- local bulk matrix versus public GEO TPM comparison;
- the complete raw-TIFF instrument-tag audit;
- reproducibility and matrix-verification workbooks.

Single-cell object:
G:/糖尿病足--疼痛--中药/6.单细胞/MMP1_single_cell_analysis11/biofsci_SingleR_annotated_MMP1_analyzed.rds
File size: 13,028,447,497 bytes. The full RDS has been found and inspected, but it
is not embedded in this ZIP or deposited by this workflow. It contains RNA counts,
LogNormalize data and scale.data layers, plus pca, harmony and umap reductions.
The supplied scripts' old H: input path must be replaced by the actual RDS path.
The original filtering thresholds, full Harmony call and SingleR reference were
not recovered from its saved history. Sample accession IDs are not donor IDs.

Bulk discovery input:
The first 1,715 gene rows and all 36,015 corresponding values exactly match the
public GSE199939 TPM matrix. The original scripts round these values for DESeq2
and a count-based VST. TPM is not raw read counts. Original manuscript results
are retained by author instruction; an appropriate reanalysis has not been run.
This is a partial file comparison, not a full-file identity assertion.

Tissue experiment:
The author clarified that DFU and control specimens were paired, with control
defined as non-ulcerated periwound tissue from the same patient. The reported
analysis remains n=3 pairs and technical triplicates. The archived PCR export has
four IDs/group and five wells/target; the exact plotted subset is unresolved.
Original unpaired P values are preserved but are not validated for the paired
design. No wells were selected or deleted to reproduce a desired P value.
The archived LC96 project differs from the reported volume/cycling conditions;
its relationship to Figure 7 remains unresolved.

All 22 supplied WB TIFFs identify an Amersham Imager 600, operation software 1.2.0.
This acquisition-software version is not the ImageJ version. Lane/specimen
mapping, exact background subtraction and quantitative intensities remain
unavailable. Original uncropped TIFFs are included without modification.

The full CTD snapshot and binary caches are not included. Existing filtered
mappings and source scripts are supplied; historical database versions/dates
remain incompletely documented. The upstream bulk-merge history is unavailable.

Every included file is listed in file_manifest.csv with its size and SHA-256.
No journal upload or public repository deposit is claimed.
