Abstract 136: Refining the molecular profile of colorectal tumors to expand prevention and treatment opportunities
Bibliographic record
Abstract
Abstract The completion of The Cancer Genome Atlas (TCGA) project for colorectal cancer (CRC) is ushering in a new phase of identifying treatment strategies tailored to the molecular profile of each person's tumor. Precision medicine approaches to cancer treatment rely on the identification of molecular profiles that can be used to identify effective therapies and can be used in a targeted sequencing setting to make treatment decisions. The initial TCGA colorectal effort included 276 samples and focused on integrating data from exome sequencing with genome-wide DNA copy number alterations (CNAs), DNA methylation, and mRNA and microRNA expression. Since then a total of 626 samples have been completed with the potential to refine CRC subtypes, identify novel mutated pathways, and further functional understanding. Such a large data set presents opportunities to identify new recurrent drug targets and to stratify patients into groups that are predictive of treatment response. However, large data sets also present substantial challenges, since hand-curation becomes intractable, while computational tools can be overwhelmed by hypermutation and copy number changes. Here we present a comprehensive molecular analysis of all 626 TCGA colorectal cancer samples, including exome sequencing, CNAs, DNA methylation, and mRNA expression. For each data type, we identified recurrently altered genes. Using MutSigCV on 525 samples yielded 27 and 87 significantly mutated genes in non-hypermutated and hypermutated samples, respectively, a substantial increase over the 15 and 17 somatically recurrently mutated genes identified using MutSig in non-hypermutated and hypermutated samples, respectively, in the previously published TCGA colorectal study. For example, PTEN, a known tumor suppressor, was not reported as significantly recurrently mutated in the initial TCGA non-hypermutated set; however, it was in the larger non-hypermutated set, demonstrating the power of a larger data set for assessing the significance and relative frequency of mutations in the context of known subtypes. In addition, we integrated the somatic mutation data, copy number data, LOH data, and hyper-methylation data to identify genes, like MLH1, that are recurrently disrupted by different mechanisms. We also considered somatic mutations that are likely gain-of-function mutations based on nonrandom clustering; and we used recurrent indels to identify loss-of-function drivers in samples positive for microsatellite instability (MSI). We further classified each sample using the previously identified subtypes – BRAF+, KRAS+, APC+, CTNNB1+ (beta-catenin+), TGFBR2/SMAD4+, PTEN+ and PIK3CA+, and R-spondin fusion positive, as well as CpG Island Methylator Phenotype (CIMP) and MSI - in order to refine the relevant molecular signatures driving CRC etiology and thereby prevention and treatment paradigms. Citation Format: Catherine S. Grasso, Eve Shinbrot, Ming Yu, Max Liesersen, Mark Chaisson, Andrew Chan, Charles Connolly, James Dai, Margaret Du, Charles Fuchs, Levi Garraway, Marios Giannakis, Tabitha Harrison, Li Hsu, Jeroen Huyghe, Jasmine Mu, Shuji Ogino, Colin Pritchard, Stephen Salipante, Wei Sun, Syed H. Zaidi, Ni Zhao, William Grady, Ben Raphael, Thomas Hudson, David Wheeler, Ulrike Peters. Refining the molecular profile of colorectal tumors to expand prevention and treatment opportunities. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 136.
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 | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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".