bioXcelerate AI at ASHG 2024: Presenting Genetic Research with ImpMap and SwitchStep

  • News and insights

bioXcelerate AI at ASHG 2024: Presenting Genetic Research with ImpMap and SwitchStep


We’re thrilled to announce that our Managing Director and Chief Scientist, Chris Foley, PhD, is representing bioXcelerate AI at this year’s American Society of Human Genetics (ASHG) Annual Meeting. Held in Denver, Colorado, the ASHG 2024 conference brings together thought leaders, researchers, and innovators from around the globe to explore the latest advancements in human genetics and genomics.

Chris Foley will present our latest work on the ImpMap and SwitchStep algorithms – novel computational tools designed to address pressing challenges in genetic fine-mapping and imputation. This research is a testament to bioXcelerate’s commitment to driving scientific discovery and innovation, made possible through collaboration with leading industry partners like GSK.

About ASHG 2024

The ASHG Annual Meeting is the world’s largest conference dedicated to human genetics and genomics. It serves as a key forum for the presentation and discussion of cutting-edge science, bringing together over 6,500 researchers, clinicians, and professionals each year. With a mix of symposia, workshops, abstract-driven sessions, and networking events, ASHG offers a unique platform for sharing research, fostering collaboration, and exploring breakthroughs in genomics that have the potential to reshape healthcare.

This year’s event is no exception, featuring a range of topics from population genetics to precision medicine. For bioXcelerate AI, ASHG 2024 offers the perfect stage to showcase our latest contributions to computational genetics, particularly through the development of tools like ImpMap and SwitchStep.

Accelerating Insights in Genetic Research

The world of genetic research increasingly relies on the processing and analysis of massive datasets to identify meaningful patterns and insights. However, this process often encounters significant computational bottlenecks. Traditional fine-mapping and imputation methods can be time-consuming, especially when handling large genomic regions or multiple datasets. To address these challenges, our team at bioXcelerate AI developed the ImpMap and SwitchStep algorithms – tools designed to dramatically improve the efficiency and accuracy of genetic analysis.

SwitchStep: Precision and Speed in Fine-Mapping

SwitchStep is our advanced fine-mapping tool that tackles one of the major challenges in genetic research: identifying causal variants within extensive genomic regions. Traditionally, fine-mapping can take days or even weeks, slowing down the pace of discovery. SwitchStep, however, reduces this processing time to minutes by utilising a high-resolution matrix inversion technique paired with a unique greedy model-based selection process. This allows researchers to pinpoint specific variants associated with traits or diseases far more quickly than before, enabling rapid target identification across large regions.

SwitchStep’s performance improvements don’t stop at speed; it also achieves up to a 99% accuracy rate in identifying causal variants, making it a highly reliable tool for genome-wide association studies (GWAS) and other research contexts where accuracy is paramount.

ImpMap: Efficient Imputation of Summary Statistics

Imputation – the process of estimating missing data points – is another vital yet challenging step in genetic analysis, especially when combining datasets for meta-analysis or to protect patient privacy. Our ImpMap algorithm addresses this by introducing a streamlined imputation method tailored for summary statistics, allowing researchers to fill in gaps without compromising data integrity or security.

Implications for Drug Discovery and Beyond

The applications of SwitchStep and ImpMap extend beyond basic research; they are particularly impactful in drug discovery, where time and precision are critical. By reducing computational delays and improving the accuracy of variant identification and data imputation, these tools accelerate the journey from raw data to actionable insights. This not only aids researchers in understanding the genetic basis of diseases but also supports pharmaceutical companies in identifying new drug targets and advancing translational research.

At bioXcelerate AI, we are committed to developing machine learning and data science tools that solve complex R&D challenges and make drug discovery more efficient. Our ImpMap and SwitchStep tools represent significant steps forward in this mission, and we’re excited to share them with the global research community at ASHG 2024.

If you wish to connect with Chris at the event, or learn more about our work, contact us here.