A genome sequencing system for universal newborn screening, diagnosis, and precision medicine for severe genetic diseases

Kingsmore, Stephen F.; Smith, Laurie D.; Kunard, Chris M.; Bainbridge, Matthew N.; Batalov, Sergey; Benson, Wendy; Blincow, Eric; Caylor, Sara; Chambers, Christina D.; Angel, Guillermo del; Dimmock, David; Ding, Yan; Ellsworth, Katarzyna A.; Feigenbaum, Annette; Frise, Erwin; Green, Robert C.; Guidugli, Lucia; Hall, Kevin P.; Hansen, Christian Holm; Hobbs, Charlotte A.; Kahn, S. D.; Kiel, Mark J.; Kraan, Lucita Van Der; Krilow, Chad; Kwon, Yong H.; Madhavrao, Lakshminarasimha; Le, Jennie; Lefebvre, Sébastien; Mardach, Rebecca; Mowrey, William R.; Oh, Danny; Owen, Mallory; Powley, George; Scharer, Gunter; Shelnutt, Seth; Tokita, Mari; Mehtalia, Shyamal S.; Oriol, Albert; Papadopoulos, Stavros; Perry, James C.; Rosales, Edwin; Sanford, Erica; Schwartz, S. J.; Tran, Duke; Reese, Martin G.; Wright, Meredith S.; Veeraraghavan, Narayanan; Wigby, Kristen; Willis, Mary; Wolen, Aaron R.; Defay, Thomas

doi:10.1016/j.ajhg.2022.08.003

Cited by 81 publications

(94 citation statements)

References 75 publications

(225 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Second, after pruning 97% (60%) of the model weights, Bonito provides 81.20% (72.66%) basecalling accuracy while using 33.33× (2.62×) smaller model using unstructured pruning (structured pruning). Third, the knee point 4 for unstructured pruning and structured pruning is at 98% and 60% where Bonito provides 65.14% and 72.66% of basecalling accuracy, respectively. Beyond the knee-point, Bonito losses its complete prediction power.…”

Section: Analyzing the State-of-the-art Basecallermentioning

confidence: 97%

A Framework for Designing Efficient Deep Learning-Based Genomic Basecallers

Singh

Alser

Khodamoradi³

et al. 2022

Preprint

View full text Add to dashboard Cite

Nanopore sequencing is a widely-used high-throughput genome sequencing technology that can sequence long fragments of a genome. Nanopore sequencing generates noisy electrical signals that need to be converted into a standard string of DNA nucleotide bases (i.e., A, C, G, T) using a computational step called basecalling. The accuracy and speed of basecalling have critical implications for every subsequent step in genome analysis. Currently, basecallers are mainly based on deep learning techniques to provide high sequencing accuracy without considering the compute demands of such tools. We observe that state-of-the-art basecallers (i.e., Guppy, Bonito) are slow, inefficient, and memory-hungry as researchers have adapted deep learning models from other domains without specialization to the basecalling purpose. Our goal is to make basecalling highly efficient and fast by building the first framework for specializing and optimizing machine learning-based basecaller. We introduce RUBICON, a framework to develop hardware-optimized basecallers. RUBICON consists of two novel machine-learning techniques that are specifically designed for basecalling. First, we introduce the quantization-aware basecalling neural architecture search (QABAS) framework to specialize the basecalling neural network architecture for a given hardware acceleration platform while jointly exploring and finding the best bit-width precision for each neural network layer. Second, we develop SkipClip, the first technique to remove all the skip connections present in modern basecallers to greatly reduce resource and storage requirements without any loss in basecalling accuracy. We demonstrate the benefits of QABAS and SkipClip by developing RUBICALL, the first hardware-optimized basecaller that performs fast and accurate basecalling. Our experimental results on state-of-the-art computing systems show that RUBICALL is a fast, accurate and hardware-friendly, mixed-precision basecaller. Compared to a highly-accurate state-of-the-art basecaller, RUBICALL provides a 16.56x speedup without losing accuracy, while also achieving a 6.88x and 2.94x reduction in neural network model size and the number of parameters, respectively. Compared to the fastest state-of-the-art basecaller, RUBICALL provides a 3.19x speedup with 2.97% higher accuracy. We show that QABAS and SkipClip can help researchers develop hardware-optimized basecallers that are superior to expert-designed models.

show abstract

Section: Analyzing the State-of-the-art Basecallermentioning

confidence: 97%

A Framework for Designing Efficient Deep Learning-Based Genomic Basecallers

Singh

Alser

Khodamoradi³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…The debate surrounding the use of NGS gene panels and exome or whole genome sequencing instead of or in addition to standard NBS methodologies has focused on the ability to accurately interpret genomic variants, costs for national healthcare systems, and ethical, legal, and social concerns ( Yang et al, 2017 ; Downie et al, 2021 ). Pilot studies are looking at diagnostic rates and comparisons of false positive and false negative results between standard NBS and genomic sequencing approaches ( Bodian et al, 2016 ; Wojcik et al, 2021 ; Kingsmore et al, 2022 ). While more detailed evaluation of the advantages and disadvantages of both methods is necessary, these initial studies suggest that genomic NBS for selected diseases would be valuable to complement ENBS programs.…”

Section: Resultsmentioning

confidence: 99%

Opportunities and challenges for newborn screening and early diagnosis of rare diseases in Latin America

Giugliani

Taucher

Hafez³

et al. 2022

Front. Genet.

View full text Add to dashboard Cite

Rare diseases (RDs) cause considerable death and disability in Latin America. Still, there is no consensus on their definition across the region. Patients with RDs face a diagnostic odyssey to find a correct diagnosis, which may last many years and creates a burden for caregivers, healthcare systems, and society. These diagnostic delays have repercussions on the health and economic burden created by RDs and continue to represent an unmet medical need. This review analyzes barriers to the widespread adoption of newborn screening (NBS) programs and early diagnostic methods for RDs in Latin America and provides recommendations to achieve this critical objective. Increasing the adoption of NBS programs and promoting early diagnosis of RDs are the first steps to improving health outcomes for patients living with RDs. A coordinated, multistakeholder effort from leaders of patient organizations, government, industry, medical societies, academia, and healthcare services is required to increase the adoption of NBS programs. Patients’ best interests should remain the guiding principle for decisions regarding NBS implementation and early diagnosis for RDs.

show abstract

“…In practice this led to an initial time‐to‐treatment goal of day of life (DOL) five. This, in turn, led to a workflow design that included heel‐prick on DOL one, screening by rWGS in 2 or 3 days, performance of the entire process in conformance with CLIA, and 1 day diagnostic interpretation and return of results accompanied by acute management guidance (Kingsmore et al, 2022). The pilot prospective clinical study will inform an assessment of the relative cost: benefit ratio of treatment on DOL 5 versus earlier and later times.A reasonable alternative scheme is that healthy babies would receive standard screening with 2‐week turnaround, with 2‐day turnaround (including shipping) reserved for the 10–15% newborns with some indication of illness that is broader than the ICU criteria used for diagnostic rWGS and that includes physician judgment.…”

Section: Discussionmentioning

confidence: 99%

“…The increasingly broad use of WGS and WES for diagnosis in affected children will continue to reclassify VUS as P, LP, or likely benign. Lastly, the artificial‐intelligence interpretation tool GEM provides a second automated method for variant prioritization (De La Vega et al, 2021; Kingsmore et al, 2022). Notwithstanding these efforts, however, BeginNGS was predicated on initial optimization of precision (positive predictive value) rather than sensitivity. Staged development.…”

Section: Discussionmentioning

confidence: 99%

Dispatches from Biotech beginning BeginNGS: Rapid newborn genome sequencing to end the diagnostic and therapeutic odyssey

Consortium¹

2022

American J of Med Genetics Pt C

Self Cite

View full text Add to dashboard Cite

In this Dispatch from Biotech, we briefly review the urgent need for extensive expansion of newborn screening (NBS) by genomic sequencing, and the reasons why early attempts had limited success. During the next decade transformative developments will continue in society and in the pharmaceutical, biotechnology, informatics, and medical sectors that enable prompt addition of genetic disorders to NBS by rapid whole genome sequencing (rWGS) upon introduction of new therapies that qualify them according to the Wilson and Jungner criteria (Wilson, J. M. G., & Jungner, G., World Health Organization. (1968). Principles and Practice of Screening for Disease.World Health Organization. Retrieved from https://apps.who.int/iris/handle/10665/ 37650). Herein we describe plans, progress, and clinical trial designs for BeginNGS (Newborn Genome Sequencing to end the diagnostic and therapeutic odyssey), a new international, pre-competitive, public-private consortium that proposes to implement a self-learning healthcare delivery system for screening all newborns for over 400 hundred genetic diseases, diagnostic confirmation, implementation of effective treatment, and acceleration of orphan drug development. We invite investigators and stakeholders worldwide to join the consortium in a prospective, multi-center, international trial of the clinical utility and cost effectiveness of BeginNGS.

show abstract

A genome sequencing system for universal newborn screening, diagnosis, and precision medicine for severe genetic diseases

Cited by 81 publications

References 75 publications

A Framework for Designing Efficient Deep Learning-Based Genomic Basecallers

A Framework for Designing Efficient Deep Learning-Based Genomic Basecallers

Opportunities and challenges for newborn screening and early diagnosis of rare diseases in Latin America

Dispatches from Biotech beginning BeginNGS: Rapid newborn genome sequencing to end the diagnostic and therapeutic odyssey

Contact Info

Product

Resources

About