Scalable nonparametric clustering with unified marker gene selection for single-cell RNA-seq data.

Clustering is commonly used in single-cell RNA-sequencing (scRNA-seq) pipelines to characterize cellular heterogeneity. However, current methods face two main limitations. First, they require user-specified heuristics which add time and complexity to bioinformatic workflows; second, they rely on post-selective differential expression analyses to identify marker genes driving cluster differences, which has been shown to be subject to inflated false discovery rates. We address these challenges by introducing nonparametric clustering of single-cell populations (NCLUSION): an infinite mixture model that leverages Bayesian sparse priors to identify marker genes while simultaneously performing clustering on single-cell expression data. NCLUSION uses a scalable variational inference algorithm to perform these analyses on datasets with up to millions of cells. By analyzing publicly available scRNA-seq studies, we demonstrate that NCLUSION (i) matches the performance of other state-of-the-art clustering techniques with significantly reduced runtime and (ii) provides statistically robust and biologically relevant transcriptomic signatures for each of the clusters it identifies. Overall, NCLUSION represents a reliable hypothesis-generating tool for understanding patterns of expression variation present in single-cell populations.

A Comprehensive Evaluation of Generalizability of Deep Learning-Based Hi-C Resolution Improvement Methods.

Hi-C is a widely used technique to study the 3D organization of the genome. Due to its high sequencing cost, most of the generated datasets are of a coarse resolution, which makes it impractical to study finer chromatin features such as Topologically Associating Domains (TADs) and chromatin loops. Multiple deep learning-based methods have recently been proposed to increase the resolution of these datasets by imputing Hi-C reads (typically called upscaling). However, the existing works evaluate these methods on either synthetically downsampled datasets, or a small subset of experimentally generated sparse Hi-C datasets, making it hard to establish their generalizability in the real-world use case. We present our framework-Hi-CY-that compares existing Hi-C resolution upscaling methods on seven experimentally generated low-resolution Hi-C datasets belonging to various levels of read sparsities originating from three cell lines on a comprehensive set of evaluation metrics. Hi-CY also includes four downstream analysis tasks, such as TAD and chromatin loops recall, to provide a thorough report on the generalizability of these methods. We observe that existing deep learning methods fail to generalize to experimentally generated sparse Hi-C datasets, showing a performance reduction of up to 57%. As a potential solution, we find that retraining deep learning-based methods with experimentally generated Hi-C datasets improves performance by up to 31%. More importantly, Hi-CY shows that even with retraining, the existing deep learning-based methods struggle to recover biological features such as chromatin loops and TADs when provided with sparse Hi-C datasets. Our study, through the Hi-CY framework, highlights the need for rigorous evaluation in the future. We identify specific avenues for improvements in the current deep learning-based Hi-C upscaling methods, including but not limited to using experimentally generated datasets for training.

Clinical validity assessment of genes frequently tested on intellectual disability/autism sequencing panels.

PURPOSE: Neurodevelopmental disorders (NDDs), such as intellectual disability (ID) and autism spectrum disorder (ASD), exhibit genetic and phenotypic heterogeneity, making them difficult to differentiate without a molecular diagnosis. The Clinical Genome Resource Intellectual Disability/Autism Gene Curation Expert Panel (GCEP) uses systematic curation to distinguish ID/ASD genes that are appropriate for clinical testing (ie, with substantial evidence supporting their relationship to disease) from those that are not. METHODS: Using the Clinical Genome Resource gene-disease validity curation framework, the ID/Autism GCEP classified genes frequently included on clinical ID/ASD testing panels as Definitive, Strong, Moderate, Limited, Disputed, Refuted, or No Known Disease Relationship. RESULTS: As of September 2021, 156 gene-disease pairs have been evaluated. Although most (75%) were determined to have definitive roles in NDDs, 22 (14%) genes evaluated had either Limited or Disputed evidence. Such genes are currently not recommended for use in clinical testing owing to the limited ability to assess the effect of identified variants. CONCLUSION: Our understanding of gene-disease relationships evolves over time; new relationships are discovered and previously-held conclusions may be questioned. Without periodic re-examination, inaccurate gene-disease claims may be perpetuated. The ID/Autism GCEP will continue to evaluate these claims to improve diagnosis and clinical care for NDDs.

Disease-specific ACMG/AMP guidelines improve sequence variant interpretation for hearing loss.

PURPOSE: The ClinGen Variant Curation Expert Panels (VCEPs) provide disease-specific rules for accurate variant interpretation. Using the hearing loss-specific American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines, the Hearing Loss VCEP (HL VCEP) illustrates the utility of expert specifications in variant interpretation. METHODS: A total of 157 variants across nine HL genes, previously submitted to ClinVar, were curated by the HL VCEP. The curation process involved collecting published and unpublished data for each variant by biocurators, followed by bimonthly meetings of an expert curation subgroup that reviewed all evidence and applied the HL-specific ACMG/AMP guidelines to reach a final classification. RESULTS: Before expert curation, 75% (117/157) of variants had single or multiple variants of uncertain significance (VUS) submissions (17/157) or had conflicting interpretations in ClinVar (100/157). After applying the HL-specific ACMG/AMP guidelines, 24% (4/17) of VUS and 69% (69/100) of discordant variants were resolved into benign (B), likely benign (LB), likely pathogenic (LP), or pathogenic (P). Overall, 70% (109/157) variants had unambiguous classifications (B, LB, LP, P). We quantify the contribution of the HL-specified ACMG/AMP codes to variant classification. CONCLUSION: Expert specification and application of the HL-specific ACMG/AMP guidelines effectively resolved discordant interpretations in ClinVar. This study highlights the utility of ClinGen VCEPs in supporting more consistent clinical variant interpretation.

Correction: ClinGen expert clinical validity curation of 164 hearing loss gene-disease pairs.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

ClinGen expert clinical validity curation of 164 hearing loss gene-disease pairs.

PURPOSE: Proper interpretation of genomic variants is critical to successful medical decision making based on genetic testing results. A fundamental prerequisite to accurate variant interpretation is the clear understanding of the clinical validity of gene-disease relationships. The Clinical Genome Resource (ClinGen) has developed a semiquantitative framework to assign clinical validity to gene-disease relationships. METHODS: The ClinGen Hearing Loss Gene Curation Expert Panel (HL GCEP) uses this framework to perform evidence-based curations of genes present on testing panels from 17 clinical laboratories in the Genetic Testing Registry. The HL GCEP curated and reviewed 142 genes and 164 gene-disease pairs, including 105 nonsyndromic and 59 syndromic forms of hearing loss. RESULTS: The final outcome included 82 Definitive (50%), 12 Strong (7%), 25 Moderate (15%), 32 Limited (20%), 10 Disputed (6%), and 3 Refuted (2%) classifications. The summary of each curation is date stamped with the HL GCEP approval, is live, and will be kept up-to-date on the ClinGen website ( https://search.clinicalgenome.org/kb/gene-validity ). CONCLUSION: This gene curation approach serves to optimize the clinical sensitivity of genetic testing while reducing the rate of uncertain or ambiguous test results caused by the interrogation of genes with insufficient evidence of a disease link.