Abstract B075: The Open Single-cell Pediatric Cancer Atlas project: Collaborative analysis of pediatric tumor data
Bibliographic record
Abstract
Abstract The Open Single-cell Pediatric Cancer Atlas (OpenScPCA) project is an open, collaborative project created to analyze publicly available data from the Single-cell Pediatric Cancer Atlas (ScPCA) Portal, with the goal of improving the quality and usability of single-cell pediatric cancer data and driving insights into pediatric cancer biology through deeper analysis of available data sets. The ScPCA Portal (https://scpca.alexslemonade.org/), developed and maintained by Alex’s Lemonade Stand Foundation (ALSF), is an open-source data resource for single-cell and single-nuclei RNA sequencing data of pediatric tumors. The ScPCA Portal currently contains summarized gene expression data for over 500 samples from a diverse set of over 50 types of cancers. All data on the portal is publicly available, uniformly processed with an open-source workflow, and ready for download in formats compatible with popular single-cell data analysis frameworks. While the data available on the ScPCA Portal is usable and useful in its current form, some limitations and many open research questions remain. For instance, while the current ScPCA processing pipeline performs some automated cell-type labeling, such methods are not always reliable. In particular, annotating malignant cells in pediatric cancer samples using automated methods is challenging because many tools and references for single-cell analysis were designed with adult healthy tissues or cancer types in mind. Expert-led cell type annotation therefore represents one opportunity to improve the data in the Portal. More broadly, the ScPCA is a unique resource for exploring open problems in applying single-cell analysis to pediatric cancer, including learning recurrent gene expression programs across samples and tumor types. To coordinate further analysis of the ScPCA data, we launched the OpenScPCA project in April 2024. The OpenScPCA project aims to engage a broad community of researchers analyzing genomic data from pediatric tumors, building off the previous success of the Open Pediatric Brain Tumor Atlas project (Shapiro et al. 2023). Our goals are to improve the utility of the ScPCA data, build consensus around the strengths and weaknesses of applying existing methods to pediatric cancer data, and test emerging methodologies. The project is conducted openly on GitHub, through which we seek to join forces with external contributors with complementary expertise while making results available in near real-time for immediate reuse and extension by the community. We invite researchers with expertise in pediatric cancer gene expression and single-cell RNA-seq analysis to participate in the OpenScPCA project. ALSF will provide contributors with support through collaboration, access to computational resources, and comprehensive documentation. We hope those participating will benefit from discovering new datasets to advance their research, gain experience with cutting-edge technologies, build their research portfolios, and join a supportive community. Get started at https://openscpca.readthedocs.io/en/latest/. Citation Format: Joshua A. Shapiro, Stephanie J. Spielman, Deepashree V. Prasad, Jennifer O'Malley, Allegra G. Hawkins, David S Mejia, Jaclyn N. Taroni. The Open Single-cell Pediatric Cancer Atlas project: Collaborative analysis of pediatric tumor data [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 B075.
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 imitationNot 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.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.012 | 0.031 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.009 | 0.011 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.005 | 0.003 |
| Open science | 0.004 | 0.010 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.031 | 0.025 |
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.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".