Tweet
The American Society of Human Genetics (ASHG) recently kicked off its 2023 Annual Meeting in Washington, DC, while celebrating the 75th anniversary of the organization. ASHG President Brendan Lee, M.D., Ph.D., Chair of the Department of Molecular and Human Genetics at Baylor College of Medicine, opened the event, reflecting on the society's evolution and its goals for the future. This year's meeting brought in over 8,000 researchers and clinicians to share and learn about the latest developments in human genetics.
The meeting underscored ASHG's commitment to advancing human genetics and genomics research, with a strong focus on improving health and wellbeing. President Lee emphasized the need for continued dialogue and action toward greater diversity, equity, and inclusion in the field, acknowledging the society's past missteps and reinforcing its commitment to integrating diverse perspectives and genomes for a comprehensive understanding of human genetics.
The event featured a wide range of exciting research, with over 3,000 abstract-driven posters and platform talks. Some of the key topics included genetics in breast cancer disparities, predictive prenatal cfDNA sequencing for maternal malignancy, and rapid genomic diagnosis in sick newborns in Japan. The Distinguished Speakers Symposium offered insights from leaders in the field who showcased opportunities in the pangenome project, cancer research, genetics education, and cytogenetics.
Aside from the great work being presented, the event was filled with focus groups, career support, valuable exhibitor talks, and parties. Attendees had the opportunity to interact with prominent figures like Eric Green, while Francis Collins could be seen rocking out on the guitar. This article cannot possibly cover every aspect of the event, but will instead highlight a few presentations and posters that demonstrate the quality and depth of this year’s meeting and the organization at large.
Platform Talks
The conference featured a long list of engaging talks, a few of which we highlight below.
DARC Side of Genetics in Cancer - Consequences of Duffy-Null in Breast Cancer Disparities
During the Distinguished Speaker Symposium, Melissa B Davis from Cornell University focused on the role of genetics in the racial disparities observed in breast cancer mortality rates, especially among women of African descent. The research from the International Center for the Study of Breast Cancer Subtypes utilized RNA sequencing and Whole Genome Sequencing to identify ancestry-specific and race-specific gene signatures in tumors. This has led to new understandings of the tumor biology of African ancestry patients, a group that is consistently underrepresented in data repositories, aiming to improve health outcomes and reduce racial disparities.
The NIH IDENTIFY Study: A Prospective Evaluation of Pregnant Women with Prenatal cfDNA Sequencing Results that Suggest Maternal Malignancy
The NIH IDENTIFY study, led by Diana W. Bianchi, Director of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, found that non-reportable or atypical results from non-invasive prenatal tests (NIPT) in asymptomatic pregnant or postpartum women might indicate an underlying malignancy. Conducted on 80 individuals, over 50% were diagnosed with cancer at their initial evaluation, which included assessments like whole-body MRI. The study highlights the importance of timely clinical evaluations following atypical NIPT results, with lymphomas and colorectal cancers being the most commonly detected malignancies in this group.
Nationwide Implementation of Rapid Genomic Diagnosis of Sick Newborn Infants in Japan
The large-scale program presented by Toshiki Takenouchi from Keio University School of Medicine, aimed to establish a comprehensive diagnostic network using online genetic counseling. This project expanded from 18 neonatal intensive care units in 8 prefectures to 99 centers in 37 prefectures, covering approximately one-third of all NICU beds in Japan. Despite the rapid expansion, the diagnostic rates remained consistent at about 48% in the pilot study and 44% in the subsequent phase, demonstrating the growing need for expedited genetic diagnosis in NICUs and the effective establishment of a nationwide diagnostic network.
Single-Cell Dissection of Bipolar Disorder
Xikun Han from MIT shared the details of a collaborative project that focused on dissecting the neurobiological mechanisms of bipolar disorder (BD) using single-cell nuclear transcriptomics (snRNA-seq) of the human prefrontal cortex. Analyzing around 500,000 single-cell transcriptomes from 88 donors, they identified 1,932 differentially expressed genes (bdDEGs) across 25 cell types and integrated these findings with genetic data. This research revealed significant overlaps in gene expression and pathways between BD and schizophrenia, which highlighted shared genetic architectures. The BD single-cell atlas created from this study provides valuable insights for understanding BD at the molecular level and identifying potential therapeutic targets.
Long-Read Sequencing of Hundreds of Human Brains Provides Insight into the Impact of Structural Variation and Methylation
The presentation by Kimberly Billingsley from the NIH covered the use of long-read sequencing to analyze structural variations and methylation in the human brain. By utilizing a new protocol with Oxford Nanopore Technologies, the researchers analyzed 222 brain samples from the North American Brain Expression Consortium, discovering 86,440 structural variants and identifying those that impact gene expression in the frontal cortex. This study emphasizes the capability of long-read sequencing to reveal disease-relevant genetic variations and is a step toward understanding the genetic basis of neurological conditions, including Alzheimer's and related dementias.
Poster Presentations
Beyond the individual talks, the event also included hundreds of amazing poster presentations. Several of our favorites are reviewed below.
Closing the Gap: Targeting Complex Medically Relevant Regions of the Human Genome
Med Mahmoud from Baylor College of Medicine shared his poster presenting a novel approach for analyzing complex genomic regions relevant to medicine. This method combines DNA capture with long-read sequencing, resulting in highly accurate genomic analysis of hard-to-map regions with extensive repeats and large duplications. The comprehensive panel, which includes genes like LPA, HLA, and SMN1/2 associated with conditions like cardiomyopathies and cancer, demonstrated higher accuracy in variant calling compared to traditional whole-genome sequencing. This approach, proven effective with 12 DNA reference samples, offers the potential for enhancing clinical applications in personalized medicine.
A Cistrome-Wide Association Study Identifies Novel Risk Loci for Renal Cell Carcinoma
Ziwei Zhang from the Dana Farber Cancer Institute presented a cistrome-wide association study (CWAS) to identify novel risk loci for Renal Cell Carcinoma (RCC). Integrating large-scale RCC GWAS data with models of regulatory element activity, Zhang and colleagues identified 13,534 regulatory elements influenced by genetic variation. The study found 12 elements significantly associated with RCC risk, including seven at known risk loci and five in previously unassociated regions. This research enhances the understanding of how genetic variations in regulatory DNA contribute to RCC risk and could lead to advances in the disease's biological understanding, risk stratification, and treatment.
Accelerated Workflows for Analyzing Long-Read Sequenced Genomes in the NIH All of Us Program
The work by Eric Dawson at NVIDIA and colleagues at Baylor College of Medicine described accelerated workflows for analyzing long-read sequenced genomes as part of the NIH All of Us Program. The initiative aims to sequence nearly 25,000 genomes using the Oxford Nanopore Promethion platform, and the presented GPU-accelerated workflows facilitate basecalling, alignment, variant calling, and phasing, improving throughput, ease of use, and accuracy. These standardized, open-source workflows are adaptable for various data scales and represent an important development in harmonizing variant calling across larger consortia.
DNA Language Models: A New Paradigm for Predicting the Impact of Human Genome Variation
Gonzalo Benegas from UC Berkeley introduced a novel DNA language model that significantly advances the prediction of human genome variation impacts. Drawing inspiration from natural language processing advancements like ChatGPT, the model achieves notable results in genome-wide variant effect prediction, outperforming existing methods with less computational resource usage. Key results include an AUROC of 92.4% in classifying ClinVar variants, superior detection of variants under purifying selection, and a 15-fold enrichment of GWAS hits in regions of extreme scores.
Comprehensive, Integrated Long-Read Genomic, Epigenomic, and Transcriptomic Characterization for Solving the Molecular Basis of Mendelian Conditions
The poster by Mitchell Vollger from the University of Washington showcased the use of long-read sequencing technologies for understanding Mendelian conditions. This involved combining HiFi sequencing, bulk MAS-Seq, and Fiber-seq for simultaneous multi-omic profiling of patient samples. This approach enabled the researchers to identify the molecular mechanisms underlying a patient's complex condition, including developmental delay and bilateral retinoblastomas, by revealing a chromosome translocation's impact on gene function and expression. Vollger’s work illustrates the potential of long-read sequencing in resolving complex genetic conditions.
As mentioned, this article only covers a fraction of the extensive research presented at the meeting. If there are any key presentations we missed, please sign in and share them with the community in the comments below. Also, don’t forget to check out part two where we review the companies and technologies that exhibited at ASHG.