Paper · S7546
Scalable multi-sample single-cell data analysis by Partition-Assisted Clustering and Multiple Alignments of Networks
PAC-MAN combines adaptive partitioning and population-network matching to address shifts between cytometry samples. Tests use simulations and prior reference datasets. Validated panels, coherent samples and downstream validation are required. Benchmark ties and losses qualify broad superiority wording; internal clade labels and a thymus percentage conflict between prose and tables. Scientific funding and author credentials do not validate unrelated UAP claims.
Read the paper at ebi.ac.ukWhat we checked
All 203 supplied XML units, 20 references and 11 structured equations were read. Eighteen main and fourteen supporting figure overviews, the main table and all S1/S2/S3 supplied values were reviewed. S3 is quoted TSV despite its .xls filename. Its 44 header fields and 228×44 data fields are complete: three additive packets read 7,995 formerly unread marker means across 205 disjoint rows, with original selected rows and semantic fields preserved. Native substring and union checks validate the actual reading records. Publisher Version 2 apparatus, caption comparisons and three prose checks do not establish full revision equivalence with the XML.
Supplied article text and selected apparatus read
How we check sources
Not checked yet
- Tiny figure and network labels, especially S13/S14, remain unreadable. Raw datasets, cited preprint, software and independent reproduction were not reviewed. Reading every supplied table value does not validate its measurements, biology or every figure glyph.
- The supplied XML has no numbered revision matching the publisher’s Version 2. A complete publisher facsimile comparison was not performed; repeated captions and related references receive no independent evidence credit.
Source limitations
- Table 1 includes ties and losses; the clade 2/8 ordering reverses within prose; thymus 84.07% conflicts with 83.73% in S3; a marrow clade 1 example has no corresponding S3 row. Causes remain unknown. Nolan’s stated roles are project administration and supervision.
- The table includes values above the displayed heatmap scale and negative or near-zero marker means; none was normalized into zero or missingness. These are transformed published aggregate profiles from prior datasets, without newly verified calibration or biological classes.
Cite this record
Ye Henry Li, Dangna Li, Nikolay Samusik, Xiaowei Wang, Leying Guan, Garry P. Nolan, Wing Hung Wong. Scalable multi-sample single-cell data analysis by Partition-Assisted Clustering and Multiple Alignments of Networks. December 27, 2017. BioParity S7546. https://bioparity.io/sources/S7546Record data (JSON)