This is the third and last entry of our blog post series Reimagining Genetic Testing: Can AI Do it All? In this entry we address how AI can transform genetic testing reporting as well as leverage genetic data throughout the patient's life, unlocking a continuous dialogue with the genome.
You can find the previous entries here:
✅ Entry 1: Understanding the significance of clinical context in refining variant interpretation.
✅ Entry 2: Navigating genomic data's vastness for enhanced interpretation.
Crafting a genetic report is a labor-intensive task requiring high expertise. Laboratory professionals navigate strict guidelines, such as rigorously applying and justifying ACMG/AMP criteria for variant classification. This often involves prioritization and literature searches to find publications that support their assessments, a process that is time-consuming and challenging. The very structure and language of the report must adhere to precise, often complex, institutional or regulatory standards. AI, particularly Large Language Models (LLMs), can significantly streamline this step by:
Automated Guideline Adherence & Text Generation: LLMs can generate draft text that follows complex stylistic and structural rules, ensuring that reports meet quality and compliance standards.
Consistent explanation of applied ACMG/AMP and other relevant guidelines: This includes formulating justifications for criteria application in a manner consistent with established best practices.
Intelligent Literature Search & Citation: Rapidly identifying and suggesting relevant scientific literature to support variant interpretations, reducing the manual burden of literature review and citation.
Amplifying Clarity: Assisting in distilling complex findings into formats readily actionable for referring clinicians and understandable during patient consultations. This ensures the initial insights are communicated with maximum impact.
By alleviating some of the most demanding aspects of report writing, AI allows experts to focus on the most nuanced aspects of interpretation, enhancing both efficiency and quality.
But what happens when, despite the very well written report, results remain inconclusive?
Results from genetic analysis are often inconclusive, with reports including Variants of Uncertain Significance (VUS). For these cases, AI can help in identifying and presenting immediate next steps, "suggesting actionable tests needed to confirm/negate inconclusive VUS findings" or highlighting potential "actionable outcomes or treatment paths."
However, this is not the only path towards higher actionability. Even for confirmed diagnoses, there are several aspects where the data driven approach used to build a report could be leveraged to generate actionable insights.
Clinical Trials recommendations could be included for information purposes, reducing the burden of patients and physicians to search for them if it is in their interest.
Similarly, pharmacogenomics for drug or dosage recommendations could be indicated, based on genetic markers found within the patient's genetic data.
Ultimately, the genetic testing report can become a navigation tool to guide next steps. Furthermore, the genetic data can become an important asset for the patient's health.
This is where the paradigm shifts. The genetic data, once sequenced, remains a rich resource. AI enables us to move beyond static interpretations by facilitating:
Dynamic Re-analysis: Imagine a physician encountering a new clinical question for a patient whose genetic data was sequenced years ago. Instead of potentially needing to re-sequence or manually sift through outdated interpretations, AI tools can enable a targeted re-interrogation of the existing dataset with this new clinical hypothesis. This could be questions like, "This patient now presents with symptom X; is there a genetic explanation within their existing data that we might not have focused on initially?".
Proactive Updates & Evolving Insights: The field of genetics is in constant flux. AI can continuously scan new research and updated databases, cross-referencing them with a patient's stored genetic data. This allows for proactive alerts to clinicians when a previously identified VUS is reclassified, or when a new gene-disease association becomes relevant to their patient's profile, effectively ensuring the genetic insights evolve alongside scientific progress.
Beyond these examples, actionable insights presented above, such as clinical trials or pharmacogenomics, could be continuously assessed depending on the patient context. This capability transforms the initial genetic test from a single diagnostic event into a continuous dialogue with the patient's genomic data.
This continuous, AI-driven approach to genetic data offers profound benefits:
For Laboratories and Insight Generators: It opens avenues for new service models, moving beyond single-report transactions to offering ongoing genomic interpretation services.
For Clinicians: It provides the tools to revisit and re-evaluate genetic information with new clinical context, increasing diagnostic yield over time and enabling more personalized treatment strategies as a patient's journey unfolds.
For Patients: As our initial concept paper envisioned, the ultimate outcome is "better understanding, higher trust, and more empowered patients." This empowerment is no longer tied to a single report but becomes a continuous journey of understanding their health, supported by evolving insights from their own genetic blueprint.
The journey towards integrating AI into genetic testing is about building a dynamic, learning ecosystem around each patient's genetic information. At Nostos Genomics, we are committed to developing AI solutions like AION that not only refine initial reporting but also unlock the immense potential of continuous genomic interrogation, supporting labs and clinicians in delivering truly personalized and evolving care.
With all this promise, can we afford not to Reimagine Genetic Testing?
Contact us!
*Nostos Genomics regularly produces webinars, white papers, and other types of content that you may find valuable.
You can unsubscribe at any time. For more information view our Privacy Policy.