Machine Learning for Biomarker Discovery in Cancer Pharmacogenomics Data
Bibliographic record
Abstract
Over the past decade, there has been an explosion in the availability of massive datasets combining drug screening with high-throughput molecular profiling in cancer model systems. These datasets have become a rich community resource which can be leveraged for biomarker discovery, in-silico validation, drug repurposing, drug method of action prediction, and to train statistical machine learning models for drug response prediction. However, this data poses unique challenges during analysis and requires methods that are robust to the noise inherent in the drug sensitivity assays. Furthermore, irreproducibility of some findings across studies strongly motivates integrative analysis across studies. Fortunately, tools have been developed implementing bioinformatics and machine learning methods designed specifically for the analysis of pre-clinical pharmacogenomics data. In this tutorial, participants will become familiar with common preclinical cancer models (such as cell-line, patient derived xenografts and organoids) and publicly available large pharmacogenomics datasets. In the hands on session, they will be introduced to the tools and packages published for analysis of these datasets, with a focus on tools written in R. Furthermore, after becoming familiar with the challenges posed by the noise in the pharmacological assays observed in high-throughput pharmacogenomics, participants will gain hands on experience using these datasets for the purpose of biomarker discovery and validation as well as building machine learning models predictive of drug response. The focus will be on translational research, validating discoveries from in vitro datasets using in vivo pharmacogenomic and clinical datasets.
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 | 0.007 | 0.023 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.003 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.007 | 0.004 |
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".