This year's American Society for Human Genetics (ASHG) conference is taking place in Denver, bringing together researchers, clinicians, and industry leaders from around the world. The meeting has a diverse lineup of talks, workshops, and poster sessions, covering topics like genetic diseases, innovations in multi-omics, and precision medicine. In this article, we’ll review some of our favorite sessions, including standout talks, poster presentations, and notable research findings. Afterwards, explore segment two, where we’ll explore the latest technology updates, cutting-edge tools, and key announcements from the industry.
Standout Talks
Charting the Landscape of Somatic Mutations: The SMaHT Consortium’s Goal for Understanding Human Tissue Variability
Tim Coorens explained how the SMaHT Network aims to catalog somatic mutations across various healthy human tissues and address challenges in mutation detection within non-diseased cells. Using advanced sequencing methods, this collaborative effort will produce a detailed reference to support studies on mutation patterns throughout the human lifespan and their implications for health and aging.
A Complete Telomere-to-Telomere Reference Panel of 6404 Human Haplotypes Improves Imputation and Phasing Accuracy
The presentation from Joseph Lalli focused on a newly developed telomere-to-telomere (T2T) reference panel of 6,404 human haplotypes that improves the accuracy of genetic imputation and phasing. The T2T panel shows reduced error rates provides a valuable resource for investigating genetic variations in previously inaccessible genomic regions and enables more accurate genetic studies.
Detection of Mosaic Structural Variation with Sniffles2
Luis Paulin presented his work on Sniffles2, a bioinformatic tool designed to detect mosaic structural variations (SVs) within sequencing data, including low-frequency SVs. Sniffles2 demonstrated accuracy in detecting SVs specific to neurodegenerative diseases and provided benchmarks in both cancer cell lines and neurodegenerative disease samples.
CellPhenoX: An eXplainable Cell-specific machine learning method to predict clinical Phenotypes using single-cell multi-omics
During the AI and machine learning sessions, Jade Young described their work on CellPhenoX. This machine learning tool analyzes single-cell multi-omics data to identify cell phenotypes predictive of clinical traits. It uses interpretable scores to differentiate phenotypes, with applications demonstrated in datasets for COVID and ulcerative colitis. CellPhenoX promises broad use in linking cell-specific data to clinical outcomes across diverse conditions.
Long-read RNA-sequencing demarcates cis- and trans-directed alternative RNA splicing
Giovanni Quinones-Valdez shared their work on the isoLASER tool that uses long-read RNA sequencing to classify alternative splicing events influenced by cis- and trans-acting factors. This approach reveals allele-specific splicing events, particularly in immune-related genes, and identifies variants linked to splicing in disease-relevant genes, improving insights into splicing mechanisms.
Poster Presentations
Integrating massive-scale GWAS and multi-omic datasets reveals disease relevant cell types and effector genes for inflammatory bowel disease
Laura Fachal shared her study that integrates large-scale GWAS and multiomic data to identify cell types and genes linked to inflammatory bowel disease (IBD). The analysis of diverse population data led to the discovery of 290 new disease signals, showing both shared and subtype-specific susceptibilities across IBD types. Mechanistic insights were also demonstrated from the study, including the role of specific genes (BCL2 and BCL2L11) associated with IBD treatment responses.
Genome wide association studies of 1,728 traits in a multi-ancestry cohort from the Michigan Genomics Initiative
Emily Bertucci-Richter analyzed a diverse cohort from the Michigan Genomics Initiative, conducting 1,728 GWAS on various traits to address their underrepresentation in genetic studies. The study effectively controlled for population stratification and replicated findings from previous studies while identifying novel associations. The approach used in this work streamlines multi-ancestry cohort analysis and preserves data from admixed individuals.
Integrative analysis of RNA-Seq, long-read DNA sequencing, and clinicopathological profiles of LINE-1 retroelement in non-small cell lung cancer
Yingshan Wang from the Ramos lab shared her work on LINE-1 retrotransposon expression in non-small cell lung cancer (NSCLC) using RNA-Seq and long-read DNA sequencing. She found elevated LINE-1 levels in lung squamous cell carcinomas compared to other NSCLC subtypes. Expression patterns were also influenced by race and age, with higher levels in African Americans and older individuals. The data suggest a potential link between LINE-1 and immune cell infiltration, particularly involving macrophages, in the tumor microenvironment.
Development of Cross-ancestry Polygenic Risk Scores for Breast Cancer in Women with African Ancestry
The research James Li presented focuses on developing cross-ancestry polygenic risk scores (PRS) for breast cancer, especially for African American women, where PRS accuracy has been limited. The study found substantial heritability estimates across breast cancer subtypes, with the best models showing strong predictive ability. The results also suggest further validation and clinical assessment for broader applicability.
Novel non-coding promoter mutations in BCL2L12 and IRF3 impacting 5% of melanoma patients
Xander Janssens studied non-coding promoter mutations in the BCL2L12 and IRF3 genes, found in a subset of melanoma patients, which impact gene expression and may influence the outcomes to immune checkpoint therapy. Functional analysis showed these mutations alter transcription factor binding, reduce promoter activity, and increase sensitivity to specific cancer therapies.
A graph-based method to characterize tandem repeats and methylation for long read genome sequencing
Mahreen Khan introduced a graph-based tool for characterizing tandem repeats (TRs) and methylation in long-read genome sequencing. Using targeted and whole-genome sequencing, the tool demonstrates high accuracy in measuring repeat size, sequence purity, and methylation. This new approach offers the potential for quicker and more precise clinical diagnoses of repeat-associated diseases.
We barely scratched the surface of all the great research presented. And with parallel presentations going on all day, we couldn't attend them all. So share some of your favorite talks or posters with is below. If you're interested in the technology and industry announcements, check out part two here!