MétaCan
Menu
← Back to cohort

Abstract B034: A transcriptional atlas provides a universal diagnostic platform for mesenchymal tumors and validation of preclinical models

2024· article· en· W4402266857 on OpenAlexaffabout
Joshua O. Nash, Pedro L. Ballestar, Scott Davidson, Astra Schwertschkow, Yael Babichev, Jodi Lees, Noa Alon, Nalan Gökgöz, Stephen Man Yu, Kyoko E. Yuki, Miranda Lorenti, Zhanqin Liu, Alaina McGoey, Famida Spatare, Bernard Castro, Kim M. Tsoi, Hagit Peretz‐Soroka, Jack Brzezinski, Anita Villani, Albiruni Razaq, Abha A. Gupta, Elizabeth G. Demicco, Joanna Przybył, Matt van de Rijn, Livia Garzia, Jay S. Wunder, Irene L. Andrulis, David Malkin, Rose Chami, Brendan C. Dickson, Rebecca A. Gladdy, Adam Shlien

Bibliographic record

VenueCancer Research · 2024
Typearticle
Languageen
FieldMedicine
TopicSarcoma Diagnosis and Treatment
Canadian institutionsMcGill UniversityUniversity Health NetworkMount Sinai HospitalLunenfeld-Tanenbaum Research InstituteHospital for Sick Children
Fundersnot available
KeywordsMesenchymal stem cellAtlas (anatomy)MedicineComputational biologyBioinformaticsCancer researchPathologyBiologyAnatomy

Abstract

fetched live from OpenAlex

Abstract Objectives: Mesenchymal neoplasms, or sarcomas, are a diverse and diagnostically challenging group including >150 histotypes, and are biased to pediatric patients, comprising 20% of pediatric solid tumor diagnoses compared to 1% in adults. A universal molecular taxonomy and classification system for sarcoma would be an invaluable tool for patient diagnosis and subtype discovery. We previously demonstrated a transcriptional pan-cancer taxonomy and classifier; however, we neither captured the full diversity of mesenchymal neoplasms nor classified preclinical models. We therefore created a second-generation, mesenchymal-specific atlas with improved diagnostic coverage, refining our ability to describe the molecular landscape of sarcoma. Methods: Fresh frozen, poly-A enriched RNA-seq datasets including mesenchymal tumors (n=24) were combined with the UCSC Treehouse Childhood Cancer Compendium. Transcriptomic data (n=14,315) were uniformly processed and all diagnostic labels were harmonized to ICD-O v3.2. To create a taxonomy, processed expression matrices were clustered one level deep and mesenchymal classes were selected and clustered to completion using established tools. To determine cluster identities, each cluster was annotated using gene expression profiles and available clinical and molecular data. We then trained a companion classifier to classify preclinical models. RNA-seq data for cell lines were obtained from the Cancer Cell Line Atlas and patient derived xenografts (PDX) from the UCSC Treehouse database. Results: All mesenchymal neoplasms (n=2,153, 82 ICD-O codes) form one cluster at the pan-cancer level, an increase of 190% in sample size and 200% in histotype diversity over our previous effort[YB1](n=1,125, 43 codes). The resulting taxonomy comprises 153 clusters over seven hierarchical levels. Out of 109 terminal nodes, 60% represent a consensus diagnosis (defined as >50% of cases classified under a single code). Of these, 50% represent novel subtypes. The remaining 40% of terminal nodes are diagnostically heterogeneous, with 25% of these containing a unifying genetic lesion, suggesting current diagnostic conventions do not capture the diversity of existing molecular entities. Sarcoma cell lines (n=54) show variable assignment to their disease of origin: several histotypes lose disease specificity, while fusion-driven histotypes conserve parental identity. Fusion-plasmid-generated models fail to recapitulate their in vivo profile. PDX’s (n=33) broadly maintain expected classification. Conclusion: RNA-seq continues to facilitate subtype discovery and disease classification in ongoing patients and preclinical models. Our results suggest a large portion of mesenchymal entities require molecular, rather than histotypic definitions. Classification of preclinical models resolves their suitability for disease modeling. The expression profiles outputted by our methods can be leveraged for future therapeutic nominations. We are expanding this effort to reach 3,000 samples this year and are open to the community to contribute. Citation Format: Joshua O. Nash, Pedro L. Ballestar, Scott Davidson, Astra Schwertschkow, Yael Babichev, Jodi Lees, Noa Alon, Nalan Gokgoz, Stephen M. Yu, Kyoko Yuki, Miranda Lorenti, Zhanqin Liu, Alaina McGoey, Famida Spatare, Bernard Castro, Kim Tsoi, Hagit Peretz-Soroka, Jack Brzezinski, Anita Villani, Albiruni Razaq, Abha Gupta, Elizabeth Demicco, Joanna Przybyl, Matt van de Rijn, Livia Garzia, Jay Wunder, Irene L. Andrulis, David Malkin, Rose Chami, Brendan C. Dickson, Rebecca A. Gladdy, Adam Shlien. A transcriptional atlas provides a universal diagnostic platform for mesenchymal tumors and validation of preclinical models [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pediatric Cancer Research; 2024 Sep 5-8; Toronto, Ontario, Canada. Philadelphia (PA): AACR; Cancer Res 2024;84(17 Suppl):Abstract nr B034.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.003
metaresearch head score (Gemma)0.004
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Bench or experimental · Consensus signal: Bench or experimental
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.010
Threshold uncertainty score0.035

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0030.004
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0020.002
Science and technology studies0.0010.001
Scholarly communication0.0020.001
Open science0.0020.002
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0100.008

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.199
GPT teacher head0.443
Teacher spread0.244 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designBench or experimental
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2024
Admission routes2
Has abstractyes

Explore more

Same venueCancer Research→Same topicSarcoma Diagnosis and Treatment→French-language works237,207→