8 December 2025 in Nature Communications
Hi-C characterizes three-dimensional chromatin organization, facilitates haplotype phasing, and enables genome-assembly scaffolding, but encounters difficulties across complex regions. By coupling chromosome conformation capture (3 C ) with PacBio H iFi long-read sequencing, here we develop a method (CiFi) that enables analysis of genomic interactions across repetitive regions. Starting with as little as 60,000 cells (sub-microgram DNA), the method produces multi-kilobasepair HiFi reads that contain multiple interacting, concatenated segments (~350 bp to 2 kbp). This multiplicity and increase in segment length versus standard short-read-based Hi-C improves read-mapping efficiency and coverage in repetitive regions and enhances haplotype phasing. CiFi pairwise interactions are largely concordant with Hi-C from a human lymphoblastoid cell line, with gains in assigning topologically associating domains across centromeres, segmental duplications, and human disease-associated genomic hotspots. As CiFi requires less input versus established methods, we apply the approach to characterize single small insects: assaying chromatin interactions across the genome from an Anopheles coluzzii mosquito and producing a chromosome-scale scaffolded assembly from a Ceratitis capitata Mediterranean fruit fly. Together, CiFi enables assessment of chromosome-scale interactions of previously recalcitrant low-complexity loci, low-input samples, and small organisms.
30 January 2025 in Autism Research
Among autistic individuals, a subphenotype of disproportionate megalencephaly (ASD‐DM) seen at three years of age is associated with co‐occurring intellectual disability and poorer prognoses later in life. However, many of the genes contributing to ASD‐DM have yet to be delineated. In this study, we identified additional ASD‐DM candidate genes with the aim to better define the genetic etiology of this subphenotype of autism. We expanded the previously studied sample size of ASD‐DM individuals ten fold by including probands from the Autism Phenome Project and Simons Simplex Collection, totaling 766 autistic individuals meeting the criteria for megalencephaly or macrocephaly and revealing 154 candidate ASD‐DM genes harboring de novo protein‐impacting variants. Our findings include 14 high confidence autism genes and seven genes previously associated with DM. Five impacted genes have previously been associated with both autism and DM, including CHD8 and PTEN. By performing functional network analysis, we expanded to additional candidate genes, including one previously implicated in ASD‐DM (PIK3CA) as well as 184 additional genes connected with ASD or DM alone. Using zebrafish, we modeled a de novo tandem duplication impacting YTHDF2, encoding an N6‐methyladenosine (m6A)‐mRNA reader, in an ASD‐DM proband. Testing zebrafish CRISPR knockdown led to reduced head/brain size, while overexpressing YTHDF2 resulted in increased head/brain size matching that of the proband. Single‐cell transcriptomes of YTHDF2 gain‐of‐function larvae point to reduced expression of Fragile‐X‐syndrome‐associated FMRP‐target genes globally and in the developing brain, providing insight into the mechanism underlying autistic phenotypes. We additionally discovered a variant impacting a different gene encoding an m6A reader, YTHDC1, in our ASD‐DM cohort. Though we highlight only two cases to date, our study provides support for the m6A‐RNA modification pathway as potentially contributing to this severe form of autism.
21 December 2022 in medRxiv
Among autistic individuals, a subphenotype with brain enlargement disproportionate to height (autism with disproportionate megalencephaly, or ASD-DM) seen at three years of age is associated with co-occurring intellectual disability and poorer prognoses later in life. However, many of the genes contributing to ASD-DM have yet to be delineated. In this study, we aim to identify additional ASD-DM associated genes to better define the genetic etiology of this subphenotype of autism. Here, we expand the previously studied sample size of ASD-DM individuals ten-fold by including probands from the Autism Phenome Project and Simons Simplex Collection, totaling 766 autistic individuals meeting the criteria for megalencephaly or macrocephaly and revealing 153 candidate ASD-DM genes harboring de novo protein-impacting variants. Our findings include thirteen high confidence autism genes and seven genes previously associated with DM. Five impacted genes have previously been associated with both autism and DM, including CHD8 and PTEN. By performing functional network analysis, we also narrowed in on additional candidate genes, including one previously implicated in ASD-DM (PIK3CA) as well as 184 additional genes previously implicated in ASD or DM alone. Using zebrafish as a model, we performed CRISPR gene editing to generate knockout animals for seven of candidate genes and assessed head-size and induced seizure activity differences. From this analysis, we identified significant morphological changes in zebrafish knockout models of two genes, ythdf2 and ryr3. While zebrafish knockout mutants model haploinsufficiency of assayed genes, we identified a de novo tandem duplication impacting YTHDF2 in an ASD-DM proband. Therefore, we also transiently overexpressed YTHDF2 by injection of in vitro transcribed human mRNA to simulate the patient-identified duplication. Following this, we observed increased head and brain size in the YTHDF2 overexpression zebrafish matching that of the proband, providing a functional link between YTHDF2 mutations and DM. Though discovery of additional mutations of ASD-DM candidate genes are required in order to fully elucidate the genetics associated with this severe form of autism, our study was able to provide support for YTHDF2 as a novel putative ASD-DM gene.