How ML can Help Tailor Immunotherapies to Cancer Patients

on YouTube ↗28m

" summary

Single-cell cytometry measures millions of cells per sample across ~40–50 proteins and a pipeline uses single-cell cytometry plus cloud compute (using AWS Batch / Nextflow / Docker) to exhaustively sample ~1,000 annotation trees. On a 27‑patient Merkel cell carcinoma trial (18 responders) an algorithm called FAUST enumerates 200–300 distinct T‑cell populations, finds two CD3+ CD8+ PD‑1+ (and CD28+ HLA‑DR+) populations that are significantly enriched in responders, and replicates on an independent melanoma dataset.

AI-written summary · about

" chapters (8)

1. 0:00Introduction and context
2. 2:35Challenges in cancer data
3. 5:23Limitations of manual analysis
4. 8:08Developing the FAST algorithm
5. 14:09Computational scaling
6. 16:08Immunotherapy for skin cancer
7. 18:37Identifying biomarkers
8. 22:14Future goals and conclusion

" related talks

• 21mFighting Cancer with AISebastian Kuss, Torsten Haferlach · DLD Munich 24
• 22mReimaging HealthcarePeter Lee · HL7 FHIR DevDays 2019 Redmond
• 29mMachine Learning Techniques for Omics Data AnalysisVesna Pajic · DSC Europe 23
• 32mHow machine learning helps cancer researchEvelina Gabasova · Strange Loop 2015
• 33mUnderstanding Molecular Complexity for Precision MedicineNatasa Przulj · From Algorithms to Discovery in Genome-Scale Biology and Medicine 2022
• 44mNext-Gen TherapiesMalin Stridh, Konstantinos Vougas · Øredev 2023
• 32mMulti-Omic Integration for Understanding DiseaseRon Shamir · Computational Challenges in Very Large-Scale 'Omics'
• 29mCancer diagnostics enabled by machines: finding and classifying single tumor....Marc Päpper · code.talks 2023