Augmenting machine learning with experimental data for drug...
summary
Scipher Medicine deploys a predictive blood test called prismaray: a 23-biomarker panel combining 10 SNPs, transcript measures and five clinical features into a single score via random forest. Validation on 175 biologic‑naive RA patients from the CORRONA registry yields PPV ≈90%, specificity ≈87% and odds ratio 6.6 for inadequate response. Stratifying therapy with prismaray raises response from ~30% to ~43% (≈40% relative gain) and models show $0.11 net PMPM savings — about $7.3M for a 1M‑member plan.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 1:50About Scipher Medicine
- 2:25Founders
- 2:54Backers
- 3:19Autoimmunity
- 4:26Autoimmunity vs oncology
- 7:07AI datadriven platform
- 8:23What is the unmet need
- 9:16Networks
- 12:34First product
- 13:15Summary
- 14:53How did you start
- 15:52Rheumatoid arthritis
- 17:23Prism Array
- 18:00Corona Registry
- 19:22Clinical Utility
- 20:59Health Economic Impact
- 21:52Summarize
- 23:09Questions
related talks
20mTranslating a Trillion Pointo Diseases of Data into Therapies
20mDiagnostic Medical application of ML
20mInnovating Clinical Research with AI and Real World Data
53mThe Algorithm for Precision Medicine
17mPrecision Medicine for All: Using Tidymodels to Validate PRS in Brazil
22mReimaging Healthcare
28mAI in healthcare - Deep learning for precision psychiatry
30mMachine Learning: A New Approach to Drug Discovery