Pharmacogenomic discovery of genetically targeted cancer therapies optimized against clinical outcomes
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Despite the clinical success of dozens of genetically targeted cancer therapies, the vast majority of patients with tumors caused by loss-of-function (LoF) mutations do not have access to these treatments. This is primarily due to the challenge of developing a drug that treats a disease caused by the absence of a protein target. The success of PARP inhibitors has solidified synthetic lethality (SL) as a means to overcome this obstacle. Recent mapping of SL networks using pooled CRISPR-Cas9 screens is a promising approach for expanding this concept to treating cancers driven by additional LoF drivers. In practice, however, translating signals from cell lines, where these screens are typically conducted, to patient outcomes remains a challenge. We developed a pharmacogenomic (PGx) approach called "Clinically Optimized Driver Associated-PGx" (CODA-PGX) that accurately predicts genetically targeted therapies with clinical-stage efficacy in specific LoF driver contexts. Using approved targeted therapies and cancer drugs with available real-world evidence and molecular data from hundreds of patients, we discovered and optimized the key screening principles predictive of efficacy and overall patient survival. In addition to establishing basic technical conventions, such as drug concentration and screening kinetics, we found that replicating the driver perturbation in the right context, as well as selecting patients where those drivers are genuine founder mutations, were key to accurate translation. We used CODA-PGX to screen a diverse collection of clinical stage drugs and report dozens of novel LoF genetically targeted opportunities; many validated in xenografts and by real-world evidence. Notable examples include treating STAG2-mutant tumors with Carboplatin, SMARCB1-mutant tumors with Oxaliplatin, and TP53BP1-mutant tumors with Etoposide or Bleomycin.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 it