Paper · S7553
Strategies for Accurate Cell Type Identification in CODEX Multiplexed Imaging Data
The study compares 20 cell-annotation pipelines using four healthy-colon tissue blocks from one donor. Agreement improves with broader labels, but both reference sets use the same images and one explicitly favors a pipeline being evaluated. The authors recommend checking expression patterns and image locations when assigning cell types. Marker counts, a worked F-score example and several caption or settings inconsistencies remain documented. These results do not establish universal clinical accuracy or a UAP connection.
Read the paper at ebi.ac.ukWhat we checked
All 121 supplied XML units and 25 references were read. Six meaningful main figures, all 20 supplement pages containing 16 figures, and all 1,365 workbook grid positions, including 507 populated cells and eight comments, were reviewed. Image expressions were inspected; the XML has no MathML. August 13, 2021 journal publication differs from September 4 PMC release. Nolan is a co-corresponding author with conceptualization, writing and funding roles.
Supplied text and scoped apparatus read
How we check sources
Not checked yet
- The original main-layout PDF, complete correction or version history, linked reference bodies, raw data and executable code were not reviewed.
- Exact unprinted plot values, every tiny gating-screen label and illustrative cell identifier were not read. A transformation-sign glyph remains ambiguous after inspection. Worksheet styles were structurally checked without an Excel layout review; duplicate figure renditions add no reading credit.
Source limitations
- The source gives 49 versus 47 antibodies, percentage outliers contrary to a blanket proximity claim, and a printed F-score inconsistent with its worked inputs. Caption references and workbook settings also conflict; no silent repairs were made.
- One donor and internally constructed references limit generalization. Industry support, patent and equity or advisory relationships are reported with their original qualifications. Government funding does not establish classified access or a programme invitation.
Cite this record
John W. Hickey, Yuqi Tan, Garry P. Nolan, Yury Goltsev. Strategies for Accurate Cell Type Identification in CODEX Multiplexed Imaging Data. August 13, 2021. BioParity S7553. https://bioparity.io/sources/S7553Record data (JSON)