Welcome to another blog post!
Genetic variants play an increasingly important role in medical diagnostics, helping clinicians identify disease risk, confirm diagnoses, predict disease progression, guide treatment decisions, and personalise patient care. Advances in sequencing technologies over recent years have meant that identifying genetic variants has never been easier1. However, interpreting these genetic variants and determining pathogenicity represents a significant challenge for clinical and research laboratories alike.
This becomes particularly pronounced in small labs that can lack the resources required for the lengthy and complex variant interpretation processes. Recently, AI-driven tools have emerged as a potential solution; demonstrating their ability to simplify workflows, enhance time- and cost-efficiency, and result in more robust, accurate genetic variant interpretations and diagnoses. In this article, we will explore the challenges of traditional genetic variant interpretation methods and investigate the potential of AI-driven tools as solutions for small laboratories.
Traditional genetic variant interpretation relies on a multi-step process to determine the pathogenicity of a variant. This approach typically follows guidelines set by the American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP) and utilises various tools and databases2,3.
.png)
1. Variant Calling: Sequencing data is aligned to a reference genome to identify where the aligned reads differ from the reference genome, identifying potential variants and creating a Variant Call Format (VCF) file for further analysis4.
2. Annotation: Identified variants are next annotated with information based on prior knowledge and evidence. This process helps to later prioritise variants for further study by predicting their biological function or potential impact on health and disease.
3. Filtering: Filtering takes into account the quality and depth of coverage of the data, as well as allele frequencies, to rule out artefacts, false positives, and common benign variants, using in-house and external curated databases5.
4. Prioritisation: Variants are then prioritised based on their likelihood of pathogenicity. This is generally done by combining a thorough review of the scientific literature with in silico tools that predict the functional impact of variants based on conservation, protein structure, or known biochemical pathways.
5. Clinical Interpretation: Finally, clinical interpretation is performed to define the likelihood of causality. This step requires the consideration of clinical, family, and functional assay data and usually involves input from highly specialised and experienced clinical experts2.
This multi-step process requires complex decision-making, manual curation, and input from bioinformatics specialists, geneticists, and clinical experts, all while navigating multiple databases and tools2. As a result, traditional variant interpretation is often challenging for smaller labs, which face difficulties in managing such resource-intensive, time-consuming tasks.
Workflow Complexity - Traditional variant interpretation involves a labour-intensive, multi-step process requiring manual curation and data entry. Due to the complexity of these workflows, laboratories often find it difficult to manage them efficiently. This complexity can lead to inefficiencies in their work and increases the chance of overlooking important information during the long and involved analysis process.
Time Constraints - Analysing large datasets from whole exome (WES) or whole genome sequencing (WGS) using manual methods can take days or weeks, delaying diagnoses and negatively impacting patients.
Limited Resources - Budget restrictions might limit access to the latest advanced tools and technologies, such as computational prediction tools. This financial constraint can further hinder the accuracy and speed of variant interpretation.
Specialised Expertise Needs - Small laboratories often have limited in-house bioinformatics or clinical genetics specialists, which makes the interpretation process challenging and limits the rate of diagnosis.
Risk of Inaccuracy - Inconsistent results from various databases and prediction tools, combined with subjective manual interpretation, can lead to interpretation inaccuracies. This variability may undermine confidence in the results, leading to cases remaining undiagnosed.
Over the past few years, AI-powered tools have revolutionised the way genetic variants are interpreted, especially in the context of rare disease diagnosis. AI tools can be applied to the variant calling, annotation, filtering, and prioritisation processes or the entire variant interpretation process6. Many AI tools have been developed based on the automatic implementation of ACMG/AMP guidelines, and since they are categorised as Class C in vitro diagnostic devices under the new In Vitro Diagnostics Regulation (IVDR), they must comply with strict regulatory requirements3,7. For small labs often facing the inherent challenges of limited resources, time constraints, and workflow complexity, AI-based platforms therefore offer a highly advantageous solution.
Simplifying and Expediting Complex Workflows - AI tools automate key steps in the variant interpretation workflow, eliminating the need for manual cross-referencing and reducing the complexity of the analysis process, enabling faster results.
Enhancing Efficiency and Diagnostic Yield - By automating tasks like annotation and prioritisation, AI platforms allow small labs to perform robust and complex analyses efficiently in less time. This frees up staff to focus on complex tasks that require their training and skill and, ultimately, results in more diagnoses.
Supporting Challenging Diagnoses - AI-powered tools excel at identifying rare or difficult-to-classify variants, using large datasets and machine learning models to improve the diagnosis of complex cases that traditional methods might miss.
Improving Diagnostic Accuracy - AI platforms increase diagnostic accuracy by leveraging high-quality data and consistent algorithms. They produce reliable and repeatable results, reducing variability and human error.
.png)
AION, Nostos Genomics’ AI-driven variant interpretation platform, provides small labs with an efficient, flexible, and scalable solution for diagnosing rare diseases. Traditional methods of variant interpretation can be time-consuming and struggle to keep pace with the volume and complexity of data generated by WES and WGS. AION addresses these challenges by using a machine-learning model trained on high-quality genetic variant data, significantly reducing the time needed for analysis.
The variant interpretation process with AION is streamlined into a simple three-step workflow, completed within minutes:
AION has been clinically validated through the Genomics England 100,000 Genomes Project, demonstrating 93.1% sensitivity in detecting causative variants. It also outperforms traditional rule-based ACMG classification by 50% in the proportion of solved cases. Within minutes, AION generates actionable reports that not only classify variants but provide clear insights into the reasoning behind each classification, helping labs make informed decisions and enhancing trust in these decisions. Additionally, AION integrates seamlessly into existing workflows, enabling labs to enhance their diagnostic efficiency and provide more timely, precise diagnoses to patients with rare genetic disorders.
AI-driven tools are transforming genetic variant interpretation, especially for small labs that struggle with the limitations of traditional methods. By automating complex workflows, reducing the need for specialised expertise, and improving diagnostic accuracy, these platforms provide a scalable, efficient solution to the challenges of genetic testing. Tools like AION streamline the entire interpretation process, allowing small labs to analyse large datasets quickly and reliably, making genetic testing more accessible.
To learn more about how Nostos Genomics and our AI-driven variant interpretation platform, AION, can support your lab in achieving more accurate, efficient variant interpretation, book a free demo with one of our genomics experts.
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.