Thursday, December 5, 2024

BLOODPAC: Webinar Held December 5, 2024, on ctDNA, Multiomics and AI

On December 5, 2024, BLOODPAC held a 2-hour, multi-speaker webinar on ctDNA, highlighting new trends in multiomics and AI.

Find BLOODPAC here.  The webinar will eventually be in a streaming archive, and a meeting report will be generated as well.  Their events archive is here.

##

Keynote speakers were:

  • Samir Hunash PhD, MD Anderson, proteins.
  • Wendy Winckler PhD, Droplet Biosciences, enhancing MRD and early detection.
  • Christopher Douville PhD, Johns Hopkins, LBx toolkit.
  • Robert Tell PhD, Tempus AI, leveraging multi-modal, multi-omic data streams
These speakers were followed by a panel, and then Q&A.

A TEMPUS slide characterized the difficulty of implementing an AI dream, which would cloud-integrate all sorts of diverse, siloed and longitundinal clinical and molecular data:

Tempus / Bloodpac / 1h9m


##

AI Corner (Chat GPT 4o)

Report on the "Beyond ctDNA: Towards the Integration of Multi-omics and Multimodal AI in Liquid Biopsy Assays" Seminar

Date: December 5, 2024
Duration: 2:00 PM - 4:00 PM ET
Host: BLOODPAC

The seminar, titled "Beyond ctDNA: Towards the Integration of Multi-omics and Multimodal AI in Liquid Biopsy Assays," convened leading experts in oncology, bioinformatics, and diagnostics to discuss advancements in multi-omics and AI applications in liquid biopsy technology. The event comprised a series of presentations, a panel discussion, and an interactive Q&A session.

Key Highlights

1. Introduction and Context

The seminar opened with remarks emphasizing the rapid advancements in liquid biopsy technologies, particularly the integration of multi-omics and multimodal data. The overarching goal was to explore how these innovations could enhance diagnostic precision and cancer treatment.

2. Presentations

2.1. Circulating Proteins for Cancer Risk Assessment
The first presentation focused on the use of circulating protein biomarkers for assessing cancer risk. The discussion included:

  • Development of multi-cancer screening panels incorporating protein markers.
  • Validation of markers like pro-surfactant B for lung cancer risk and its integration with imaging findings.
  • A trajectory-based approach to cancer risk detection, showcasing its potential to identify high-risk patients earlier.

2.2. Challenges Beyond ctDNA
This session addressed the limitations of circulating tumor DNA (ctDNA) in low-tumor-burden settings and proposed strategies to overcome these challenges:

  • Use of complementary analytes such as RNA, metabolites, and proteins.
  • Examples of assays improving sensitivity for early detection and post-operative monitoring, including lymphatic fluid analysis for head and neck cancers.

2.3. Analytical Toolkit for Liquid Biopsy Development
A detailed discussion on the statistical and computational methods used to evaluate candidate biomarkers, including:

  • Tools for detecting rare mutations, aneuploidy, and fragmentation patterns.
  • Strategies to mitigate overfitting in AI-based liquid biopsy models.

2.4. Multi-modal and Multi-omics Data Streams in Diagnostics
The final presentation highlighted the integration of multiple data streams:

  • Leveraging analytes like ctDNA, RNA, and proteomics with clinical data.
  • Challenges in managing and analyzing large-scale multimodal datasets within clinical timeframes.

3. Panel Discussion

The panel featured experts in bioinformatics, diagnostics, and AI, who discussed:

  • The future of multi-omics in personalized medicine.
  • Strategies for standardizing pre-analytics and data integration.
  • Insights into regulatory and logistical challenges in deploying these advanced diagnostics.

4. Q&A Session

The event concluded with a 20-minute Q&A session, during which participants posed questions about the practical applications and challenges of integrating multi-omics and AI into clinical workflows. Key topics included:

  • Enhancing biomarker validation across diverse populations.
  • Addressing pre-analytical variability in novel assays.
  • The role of collaboration between research institutions and industry in accelerating innovation.

Conclusions

The seminar underscored the transformative potential of combining multi-omics with AI to redefine liquid biopsy technologies. Presenters and panelists stressed the importance of interdisciplinary collaboration, robust data validation, and streamlined workflows to bring these innovations from research to clinical practice. The discussions reflected optimism for the future of precision oncology and its ability to improve patient outcomes through early detection and personalized care.