can we give notebookML a collaborative try?
Review
Cancer Cell - 2022 Oct 10;40(10):1095-1110. doi: 10.1016/j.ccell.2022.09.012.
Jana Lipkova, Richard J Chen, Bowen Chen, Ming Y Lu, Matteo Barbieri, Daniel Shao, Anurag J Vaidya, Chengkuan Chen, Luoting Zhuang, Drew F K Williamson, Muhammad Shaban, Tiffany Y Chen, Faisal Mahmood
https://notebooklm.google.com/notebook/93516bd1-62f6-4d0a-93f1-146c31140128/audio
Loose bits
http://bit.ly/tcgascopeGBM
https://chirp.od.nih.gov (register when you can!)
https://notebooklm.google - lets try it with, for example, https://pubmed.ncbi.nlm.nih.gov/36113157 (this afternoon's speaker)
Do we need an agent for TCGA case IDs ... https://portal.gdc.cancer.gov/cases/9da462b0-93c2-4305-89f6-7199a30399a7
challenge
Can we still find the same information via cBio's API?
or GDC API?
genAI Dashboards (something missing?):
Ultimately, the goal is to process new observations
ESE
ARPA-H
tcgaPath
Tensorflow projector - https://bit.ly/tcgareps
- note interactive labelling and other functionalities
- ...
...
https://notebooklm.google.com/notebook/29eb533b-03da-4e54-8b63-5db44e472803/audio
...
BIOSTATISTICS BRANCH SEMINAR SERIES PRESENTS
BB seminar:
Sandra Safo, Ph.D.
Associate Professor
Division of Biostatistics & Health Data Science, University of Minnesota
Date: Wednesday, January 15th, 2025
Time: 11:30 am to 12:30 pm (EST.)
Location: NCI, Shady Grove, 7E032/034 & Join Via WebEx
Meeting number: 2311 556 0768
Password: BBsem01.15
Join by phone: 1-650-479-3207 Call-in toll number (US/Canada)
Access code: 2311 556 0768
Title: Explainable Nonlinear Multi-Task Learning for Multimodal Data Integration
Abstract:
Statistical and machine learning methods for multimodal data integration have attracted significant attention in recent research, driven by their potential to extract meaningful insights from complex, multi-source or multimodal data. In this talk, I will present innovative supervised and unsupervised nonlinear methods for data integration, with a focus on their applications in biomarker discovery and data reconstruction. Our approach combines the flexibility of deep learning with the statistical advantages of both data-driven and knowledge-guided feature selection to capture nonlinear relationships across multimodal data. We uncover shared and data-specific low-dimensional representations that are linked to multiple types of outcomes or multiple tasks, enabling a unified framework for both data integration and multi-task learning, and fostering more robust and interpretable models. Through extensive simulation studies and real-world applications, we demonstrate that our methods outperform a variety of linear and nonlinear alternatives, particularly in small-sample settings. This makes them particularly valuable in biomedical research, where data are often scarce but critical insights are needed for advancing scientific understanding.
**The mission of the Biostatistics Branch (BB) is to be an outstanding biostatistics unit that can contribute to the understanding of cancer etiology and to improve public health by the development and application of quantitative methods. The BB Investigators develop statistical methods and data resources to strengthen observational studies, intervention trials, and laboratory investigations of cancer.**