Abstract 5340: Bioinformatic analyses approaches for personalized oncogenomics
Bibliographic record
Abstract
Abstract The Personalized Oncogenomics initiative at the British Columbia Cancer Agency aims to identify tumor-specific therapeutic targets in cancer patients with late stage disease who have failed standard therapy. Comprehensive profiling of individual patients' tumor(s) at the DNA and RNA level allows for characterization of altered pathways and hence identification of therapeutics designed to specifically target them. Data for each individual study included whole genome and transcriptome sequence of the fresh biopsy and whole genome sequence of patient's blood. When tissues were available, transcriptome of a matched normal sample and genome of the commonly formalin-fixed paraffin-embedded primary tumor were also sequenced. All sequencing experiments were performed on Illumina machines. Genomic data was examined, depending on the case, for germline and/or somatic mutations. These included single nucleotide variants, small insertions and deletions and copy number variations. All sequence data were assembled de novo in order to identify rearrangements causing gene fusions; transcriptome data also revealed allelic expression of variants and provided a profile for the entire transcribed genome. Differential abundance estimation was run against a rich repository of publicly available data from The Cancer Genome Atlas project and transcriptome datasets available in-house. The variants and pathways were then mapped to drug databases as well as clinical trial records. This was followed by an extensive literature search for evidence of drug combinations, drug-drug interactions and efficacy of a drug for a particular cancer type, especially those not recognized as the approved disease group for the drug under consideration. The project has sequenced 50 patients; the average length of time between acquiring the biopsy and delivering a report to clinical oncologists was 37 days. Bioinformatic analysis of the sequence data led to identification of informative or actionable targets in up to 80% of cases. The findings were not restricted to target identification but also led to change of diagnosis, treatment and characterization of tumor evolution. De novo assembly of the data in a non-small cell lung cancer patient led to the identification of the well-characterized EML4-ALK oncogenic fusion which had been missed through the use of clinically approved fluorescence in situ hybridization test. Analysis of two separate malignant masses in another patient revealed two divergent and unique tumors. Two different therapeutic were prescribed in order to target these; this led to the disappearance of both tumors and disease stabilization for 7 months. Through the design of an efficient and automated bioinformatics pipeline, individual patient's tumor specimen(s) were profiled in a clinically relevant time frame. This in turn enabled the delivery of targeted therapies and disease stabilization in patients who had no remaining standard therapeutic options. Citation Format: Katayoon Kasaian, Yaoqing Shen, Sreeja Leelakumari, Peter Eirew, Yvonne Y. Li, Erin Pleasance, Richard Corbett, Karen L. Mungall, Jacquie Schein, Andrew J. Mungall, Yongjun Zhao, Richard A. Moore, Stephen Yip, Karen Gelmon, Howard Lim, Daniel Renouf, Robyn Roscoe, Yussanne Ma, Marco A. Marra, Janessa Laskin, Steven JM Jones. Bioinformatic analyses approaches for personalized oncogenomics. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 5340. doi:10.1158/1538-7445.AM2014-5340
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 | 0.006 | 0.010 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.004 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.004 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.011 | 0.006 |
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".