Highly accurate long reads are crucial for realizing the potential of biodiversity genomics.

BACKGROUND: Generating the most contiguous, accurate genome assemblies given available sequencing technologies is a long-standing challenge in genome science. With the rise of long-read sequencing, assembly challenges have shifted from merely increasing contiguity to correctly assembling complex, repetitive regions of interest, ideally in a phased manner. At present, researchers largely choose between two types of long read data: longer, but less accurate sequences, or highly accurate, but shorter reads (i.e., >Q20 or 99% accurate). To better understand how these types of long-read data as well as scale of data (i.e., mean length and sequencing depth) influence genome assembly outcomes, we compared genome assemblies for a caddisfly, Hesperophylax magnus, generated with longer, but less accurate, Oxford Nanopore (ONT) R9.4.1 and highly accurate PacBio HiFi (HiFi) data. Next, we expanded this comparison to consider the influence of highly accurate long-read sequence data on genome assemblies across 6750 plant and animal genomes. For this broader comparison, we used HiFi data as a surrogate for highly accurate long-reads broadly as we could identify when they were used from GenBank metadata. RESULTS: HiFi reads outperformed ONT reads in all assembly metrics tested for the caddisfly data set and allowed for accurate assembly of the repetitive ~ 20 Kb H-fibroin gene. Across plants and animals, genome assemblies that incorporated HiFi reads were also more contiguous. For plants, the average HiFi assembly was 501% more contiguous (mean contig N50 = 20.5 Mb) than those generated with any other long-read data (mean contig N50 = 4.1 Mb). For animals, HiFi assemblies were 226% more contiguous (mean contig N50 = 20.9 Mb) versus other long-read assemblies (mean contig N50 = 9.3 Mb). In plants, we also found limited evidence that HiFi may offer a unique solution for overcoming genomic complexity that scales with assembly size. CONCLUSIONS: Highly accurate long-reads generated with HiFi or analogous technologies represent a key tool for maximizing genome assembly quality for a wide swath of plants and animals. This finding is particularly important when resources only allow for one type of sequencing data to be generated. Ultimately, to realize the promise of biodiversity genomics, we call for greater uptake of highly accurate long-reads in future studies.

Cross-site comparison of ribosomal depletion kits for Illumina RNAseq library construction.

BACKGROUND: Ribosomal RNA (rRNA) comprises at least 90% of total RNA extracted from mammalian tissue or cell line samples. Informative transcriptional profiling using massively parallel sequencing technologies requires either enrichment of mature poly-adenylated transcripts or targeted depletion of the rRNA fraction. The latter method is of particular interest because it is compatible with degraded samples such as those extracted from FFPE and also captures transcripts that are not poly-adenylated such as some non-coding RNAs. Here we provide a cross-site study that evaluates the performance of ribosomal RNA removal kits from Illumina, Takara/Clontech, Kapa Biosystems, Lexogen, New England Biolabs and Qiagen on intact and degraded RNA samples. RESULTS: We find that all of the kits are capable of performing significant ribosomal depletion, though there are differences in their ease of use. All kits were able to remove ribosomal RNA to below 20% with intact RNA and identify ~ 14,000 protein coding genes from the Universal Human Reference RNA sample at >1FPKM. Analysis of differentially detected genes between kits suggests that transcript length may be a key factor in library production efficiency. CONCLUSIONS: These results provide a roadmap for labs on the strengths of each of these methods and how best to utilize them.

Chromosome-Level Genome Assembly and Annotation of a Periodical Cicada Species: Magicicada septendecula.

We present a high-quality assembly and annotation of the periodical cicada species, Magicicada septendecula (Hemiptera: Auchenorrhyncha: Cicadidae). Periodical cicadas have a significant ecological impact, serving as a food source for many mammals, reptiles, and birds. Magicicada are well known for their massive emergences of 1 to 3 species that appear in different locations in the eastern United States nearly every year. These year classes ("broods") emerge dependably every 13 or 17 yr in a given location. Recently, it has become clear that 4-yr early or late emergences of a sizeable portion of a population are an important part of the history of brood formation; however, the biological mechanisms by which they track the passage of time remain a mystery. Using PacBio HiFi reads in conjunction with Hi-C proximity ligation data, we have assembled and annotated the first whole genome for a periodical cicada, an important resource for future phylogenetic and comparative genomic analysis. This also represents the first quality genome assembly and annotation for the Hemipteran superfamily Cicadoidea. With a scaffold N50 of 518.9 Mb and a complete BUSCO score of 96.7%, we are confident that this assembly will serve as a vital resource toward uncovering the genomic basis of periodical cicadas' long, synchronized life cycles and will provide a robust framework for further investigations into these insects.

Beethoven, a mouse model for dominant, progressive hearing loss DFNA36.

Despite recent progress in identifying genes underlying deafness, there are still relatively few mouse models of specific forms of human deafness. Here we describe the phenotype of the Beethoven (Bth) mouse mutant and a missense mutation in Tmc1 (transmembrane cochlear-expressed gene 1). Progressive hearing loss (DFNA36) and profound congenital deafness (DFNB7/B11) are caused by dominant and recessive mutations of the human ortholog, TMC1 (ref. 1), for which Bth and deafness (dn) are mouse models, respectively.

Dominant and recessive deafness caused by mutations of a novel gene, TMC1, required for cochlear hair-cell function.

Positional cloning of hereditary deafness genes is a direct approach to identify molecules and mechanisms underlying auditory function. Here we report a locus for dominant deafness, DFNA36, which maps to human chromosome 9q13-21 in a region overlapping the DFNB7/B11 locus for recessive deafness. We identified eight mutations in a new gene, transmembrane cochlear-expressed gene 1 (TMC1), in a DFNA36 family and eleven DFNB7/B11 families. We detected a 1.6-kb genomic deletion encompassing exon 14 of Tmc1 in the recessive deafness (dn) mouse mutant, which lacks auditory responses and has hair-cell degeneration. TMC1 and TMC2 on chromosome 20p13 are members of a gene family predicted to encode transmembrane proteins. Tmc1 mRNA is expressed in hair cells of the postnatal mouse cochlea and vestibular end organs and is required for normal function of cochlear hair cells.

Noncoding mutations of HGF are associated with nonsyndromic hearing loss, DFNB39.

A gene causing autosomal-recessive, nonsyndromic hearing loss, DFNB39, was previously mapped to an 18 Mb interval on chromosome 7q11.22-q21.12. We mapped an additional 40 consanguineous families segregating nonsyndromic hearing loss to the DFNB39 locus and refined the obligate interval to 1.2 Mb. The coding regions of all genes in this interval were sequenced, and no missense, nonsense, or frameshift mutations were found. We sequenced the noncoding sequences of genes, as well as noncoding genes, and found three mutations clustered in intron 4 and exon 5 in the hepatocyte growth factor gene (HGF). Two intron 4 deletions occur in a highly conserved sequence that is part of the 3' untranslated region of a previously undescribed short isoform of HGF. The third mutation is a silent substitution, and we demonstrate that it affects splicing in vitro. HGF is involved in a wide variety of signaling pathways in many different tissues, yet these putative regulatory mutations cause a surprisingly specific phenotype, which is nonsydromic hearing loss. Two mouse models of Hgf dysregulation, one in which an Hgf transgene is ubiquitously overexpressed and the other a conditional knockout that deletes Hgf from a limited number of tissues, including the cochlea, result in deafness. Overexpression of HGF is associated with progressive degeneration of outer hair cells in the cochlea, whereas cochlear deletion of Hgf is associated with more general dysplasia.

Mutation spectrum of MYO7A and evaluation of a novel nonsyndromic deafness DFNB2 allele with residual function.

Recessive mutations of MYO7A, encoding unconventional myosin VIIA, can cause either a deaf-blindness syndrome (type 1 Usher syndrome; USH1B) or nonsyndromic deafness (DFNB2). In our study, deafness segregating as a recessive trait in 24 consanguineous families showed linkage to markers for the DFNB2/USH1B locus on chromosome 11q13.5. A total of 23 of these families segregate USH1 due to 17 homozygous mutant MYO7A alleles, of which 14 are novel. One family segregated nonsyndromic hearing loss DFNB2 due to a novel three-nucleotide deletion in an exon of MYO7A (p.E1716del) encoding a region of the tail domain. We hypothesized that DFNB2 alleles of MYO7A have residual myosin VIIA. To address this question we investigated the effects of several mutant alleles by making green fluorescent protein (GFP) tagged cDNA expression constructs containing engineered mutations of mouse Myo7a at codons equivalent to pathogenic USH1B and DFNB2 alleles of human MYO7A. We show that in transfected mouse hair cells an USH1B mutant GFP-myosin VIIa does not localize properly to inner ear hair cell stereocilia. However, a GFP-myosin VIIa protein engineered to have an equivalent DFNB2 mutation to p.E1716del localizes correctly in transfected mouse hair cells. This finding is consistent with the hypothesis that p.E1716del causes a less severe phenotype (DFNB2) than the USH1B-associated alleles because the resulting protein retains some degree of normal function.

Modifier variant of METTL13 suppresses human GAB1-associated profound deafness.

A modifier variant can abrogate the risk of a monogenic disorder. DFNM1 is a locus on chromosome 1 encoding a dominant suppressor of human DFNB26 recessive, profound deafness. Here, we report that DFNB26 is associated with a substitution (p.Gly116Glu) in the pleckstrin homology domain of GRB2-associated binding protein 1 (GAB1), an essential scaffold in the MET proto-oncogene, receptor tyrosine kinase/HGF (MET/HGF) pathway. A dominant substitution (p.Arg544Gln) of METTL13, encoding a predicted methyltransferase, is the DFNM1 suppressor of GAB1-associated deafness. In zebrafish, human METTL13 mRNA harboring the modifier allele rescued the GAB1-associated morphant phenotype. In mice, GAB1 and METTL13 colocalized in auditory sensory neurons, and METTL13 coimmunoprecipitated with GAB1 and SPRY2, indicating at least a tripartite complex. Expression of MET-signaling genes in human lymphoblastoid cells of individuals homozygous for p.Gly116Glu GAB1 revealed dysregulation of HGF, MET, SHP2, and SPRY2, all of which have reported variants associated with deafness. However, SPRY2 was not dysregulated in normal-hearing humans homozygous for both the GAB1 DFNB26 deafness variant and the dominant METTL13 deafness suppressor, indicating a plausible mechanism of suppression. Identification of METTL13-based modification of MET signaling offers a potential therapeutic strategy for a wide range of associated hearing disorders. Furthermore, MET signaling is essential for diverse functions in many tissues including the inner ear. Therefore, identification of the modifier of MET signaling is likely to have broad clinical implications.

The tip-link antigen, a protein associated with the transduction complex of sensory hair cells, is protocadherin-15.

Sound and acceleration are detected by hair bundles, mechanosensory structures located at the apical pole of hair cells in the inner ear. The different elements of the hair bundle, the stereocilia and a kinocilium, are interconnected by a variety of link types. One of these links, the tip link, connects the top of a shorter stereocilium with the lateral membrane of an adjacent taller stereocilium and may gate the mechanotransducer channel of the hair cell. Mass spectrometric and Western blot analyses identify the tip-link antigen, a hitherto unidentified antigen specifically associated with the tip and kinocilial links of sensory hair bundles in the inner ear and the ciliary calyx of photoreceptors in the eye, as an avian ortholog of human protocadherin-15, a product of the gene for the deaf/blindness Usher syndrome type 1F/DFNB23 locus. Multiple protocadherin-15 transcripts are shown to be expressed in the mouse inner ear, and these define four major isoform classes, two with entirely novel, previously unidentified cytoplasmic domains. Antibodies to the three cytoplasmic domain-containing isoform classes reveal that each has a different spatiotemporal expression pattern in the developing and mature inner ear. Two isoforms are distributed in a manner compatible for association with the tip-link complex. An isoform located at the tips of stereocilia is sensitive to calcium chelation and proteolysis with subtilisin and reappears at the tips of stereocilia as transduction recovers after the removal of calcium chelators. Protocadherin-15 is therefore associated with the tip-link complex and may be an integral component of this structure and/or required for its formation.

Characterization of a new full length TMPRSS3 isoform and identification of mutant alleles responsible for nonsyndromic recessive deafness in Newfoundland and Pakistan.

BACKGROUND: Mutant alleles of TMPRSS3 are associated with nonsyndromic recessive deafness (DFNB8/B10). TMPRSS3 encodes a predicted secreted serine protease, although the deduced amino acid sequence has no signal peptide. In this study, we searched for mutant alleles of TMPRSS3 in families from Pakistan and Newfoundland with recessive deafness co-segregating with DFNB8/B10 linked haplotypes and also more thoroughly characterized the genomic structure of TMPRSS3. METHODS: We enrolled families segregating recessive hearing loss from Pakistan and Newfoundland. Microsatellite markers flanking the TMPRSS3 locus were used for linkage analysis. DNA samples from participating individuals were sequenced for TMPRSS3. The structure of TMPRSS3 was characterized bioinformatically and experimentally by sequencing novel cDNA clones of TMPRSS3. RESULTS: We identified mutations in TMPRSS3 in four Pakistani families with recessive, nonsyndromic congenital deafness. We also identified two recessive mutations, one of which is novel, of TMPRSS3 segregating in a six-generation extended family from Newfoundland. The spectrum of TMPRSS3 mutations is reviewed in the context of a genotype-phenotype correlation. Our study also revealed a longer isoform of TMPRSS3 with a hitherto unidentified exon encoding a signal peptide, which is expressed in several tissues. CONCLUSION: Mutations of TMPRSS3 contribute to hearing loss in many communities worldwide and account for 1.8% (8 of 449) of Pakistani families segregating congenital deafness as an autosomal recessive trait. The newly identified TMPRSS3 isoform e will be helpful in the functional characterization of the full length protein.

Targeted amplicon sequencing (TAS): a scalable next-gen approach to multilocus, multitaxa phylogenetics.

Next-gen sequencing technologies have revolutionized data collection in genetic studies and advanced genome biology to novel frontiers. However, to date, next-gen technologies have been used principally for whole genome sequencing and transcriptome sequencing. Yet many questions in population genetics and systematics rely on sequencing specific genes of known function or diversity levels. Here, we describe a targeted amplicon sequencing (TAS) approach capitalizing on next-gen capacity to sequence large numbers of targeted gene regions from a large number of samples. Our TAS approach is easily scalable, simple in execution, neither time-nor labor-intensive, relatively inexpensive, and can be applied to a broad diversity of organisms and/or genes. Our TAS approach includes a bioinformatic application, BarcodeCrucher, to take raw next-gen sequence reads and perform quality control checks and convert the data into FASTA format organized by gene and sample, ready for phylogenetic analyses. We demonstrate our approach by sequencing targeted genes of known phylogenetic utility to estimate a phylogeny for the Pancrustacea. We generated data from 44 taxa using 68 different 10-bp multiplexing identifiers. The overall quality of data produced was robust and was informative for phylogeny estimation. The potential for this method to produce copious amounts of data from a single 454 plate (e.g., 325 taxa for 24 loci) significantly reduces sequencing expenses incurred from traditional Sanger sequencing. We further discuss the advantages and disadvantages of this method, while offering suggestions to enhance the approach.

Profound, prelingual nonsyndromic deafness maps to chromosome 10q21 and is caused by a novel missense mutation in the Usher syndrome type IF gene PCDH15.

We studied a consanguineous family (Family A) from the island of Newfoundland with an autosomal recessive form of prelingual, profound, nonsyndromic sensorineural hearing loss. A genome-wide scan mapped the deafness trait to 10q21-22 (max LOD score of 4.0; D10S196) and fine mapping revealed a 16 Mb ancestral haplotype in deaf relatives. The PCDH15 gene was mapped within the critical region and was an interesting candidate because truncating mutations cause Usher syndrome type IF (USH1F) and two missense mutations have been previously associated with isolated deafness (DFNB23). Sequencing of the PCDH15 gene revealed 33 sequencing variants. Three of these variants were homozygous exclusively in deaf siblings but only one of them was not seen in ethnically matched controls. This novel c.1583 T>A transversion predicts an amino-acid substitution of a valine with an aspartic acid at codon 528 (V528D). Like the two DFNB23 mutations, the V528D mutation in Family A occurs in a highly conserved extracellular cadherin (EC) domain of PCDH15 and is predicted to be more deleterious than the previously identified DFNB23 missense mutations (R134G and G262D). Physical assessment, vestibular and visual function testing in deaf adults ruled out syndromic deafness because of Usher syndrome. This study validates the DFNB23 designation and supports the hypothesis that missense mutations in conserved motifs of PCDH15 cause nonsyndromic hearing loss. This emerging genotype-phenotype correlation in USH1F is similar to that in several other USH1 genes and cautions against a prognosis of a dual sensory loss in deaf children found to be homozygous for hypomorphic mutations at the USH1F locus.

Mutations in TRIOBP, which encodes a putative cytoskeletal-organizing protein, are associated with nonsyndromic recessive deafness.

In seven families, six different mutant alleles of TRIOBP on chromosome 22q13 cosegregate with autosomal recessive nonsyndromic deafness. These alleles include four nonsense (Q297X, R788X, R1068X, and R1117X) and two frameshift (D1069fsX1082 and R1078fsX1083) mutations, all located in exon 6 of TRIOBP. There are several alternative splice isoforms of this gene, the longest of which, TRIOBP-6, comprises 23 exons. The linkage interval for the deafness segregating in these families includes DFNB28. Genetic heterogeneity at this locus is suggested by three additional families that show significant evidence of linkage of deafness to markers on chromosome 22q13 but that apparently have no mutations in the TRIOBP gene.

DFNB48, a new nonsyndromic recessive deafness locus, maps to chromosome 15q23-q25.1.

Nonsyndromic deafness locus (DFNB48) segregating as an autosomal recessive trait has been mapped to the long arm of chromosome 15 in bands q23-q25.1 in five large Pakistani families. The deafness phenotype in one of these five families (PKDF245) is linked to D15S1005 with a lod score of 8.6 at theta=0, and there is a critical linkage interval of approximately 7 cM on the Marshfield human genetic map, bounded by microsatellite markers D15S216 (70.73 cM) and D15S1041 (77.69 cM). MYO9A, NR2E3, BBS4, and TMC3 are among the candidate genes in the DFNB48 region. The identification of another novel nonsyndromic recessive deafness locus demonstrates the high degree of locus heterogeneity for hearing impairment, particularly in the Pakistani population.

A new locus for nonsyndromic deafness DFNB49 maps to chromosome 5q12.3-q14.1.

Cosegregation of markers on chromosome 5q12.3-q14.1 with profound congenital deafness in two Pakistani families (PKDF041 and PKDF141) defines a new recessive deafness locus, DFNB49. A maximum two-point lod score of 4.44 and 5.94 at recombination fraction theta=0 was obtained for markers D5S2055 and D5S424 in families PKDF041 and PKDF141, respectively. Haplotype analysis revealed an 11 cM linkage region flanked by markers D5S647 (74.07 cM) and D5S1501 (85.25 cM). Candidate deafness genes in this region include SLC30A5, OCLN, GTF2H2, and BTF3, encoding solute carrier family 30 (zinc transporter) member 5, occludin, RNA polymerase II transcription initiation factor, and basic transcription factor 3, respectively. Sequence analysis of the coding exons of SLC30A5 in DNA samples from two affected individuals of families PKDF041 and PKDF141 revealed no mutation. The mapping of DFNB49 further confirms the heterogeneity underlying autosomal recessive forms of nonsyndromic deafness.

DFNB3, spectrum of MYO15A recessive mutant alleles and an emerging genotype-phenotype correlation.

We have now identified seven MYO15A mutations that cause congenital profound neurosensory hearing loss and a possible hypomorphic allele of MYO15A associated with moderately-severe hearing loss in 1 of 8 SMS patients. Because myosin XVA is encoded by 66 exons, screening for mutations in hearing-impaired individuals is expensive and labor-intensive in comparison to a screen for mutations in GJB2 (Cx26), for example, which has only a single protein coding exon. Among consanguineous families segregating profound, congenital hearing loss from Pakistan, approximately 10% are consistent with linkage to DFNB3 (11 of 112 DFNB families). In one-half of these DFNB3 families, we found a homozygous mutation in 1 of the 66 exons of MYO15A [25]. This suggests that mutations of MYO15A are responsible for at least 5% of recessively inherited, profound hearing loss in Pakistan. However, without the benefit of a pre-screen for linkage to DFNB3, it will be a challenge to determine the extent to which mutations of MYO15A contribute to hereditary hearing loss among isolated cases and small families in other populations.

Nonsyndromic recessive deafness DFNB18 and Usher syndrome type IC are allelic mutations of USHIC.

Human chromosome 11 harbors two Usher type I loci, USHIB and USHIC, which encode myosin VIIA and harmonin, respectively. The USHIC locus overlaps the reported critical interval for nonsyndromic deafness locus DFNB18. We found an IVS12+5G-->C mutation in the USHIC gene, which is associated with nonsyndromic recessive deafness ( DFNB18) segregating in the original family, S-11/12. No other disease-associated mutation was found in the other 27 exons or in the intron-exon boundaries, and the IVS12+5G-->C mutation was not present in 200 representative unaffected individuals ascertained from the same area of India. An exon-trapping assay with a construct harboring IVS12+5G-->C generated wildtype spliced mRNA having exons 11 and 12 and mRNA that skipped exon 12. We conclude that mutations of USHIC can cause both Usher syndrome type IC and nonsyndromic recessive deafness DFNB18.

PCDH15 is expressed in the neurosensory epithelium of the eye and ear and mutant alleles are responsible for both USH1F and DFNB23.

Recessive splice site and nonsense mutations of PCDH15, encoding protocadherin 15, are known to cause deafness and retinitis pigmentosa in Usher syndrome type 1F (USH1F). Here we report that non-syndromic recessive hearing loss (DFNB23) is caused by missense mutations of PCDH15. This suggests a genotype-phenotype correlation in which hypomorphic alleles cause non-syndromic hearing loss, while more severe mutations of this gene result in USH1F. We localized protocadherin 15 to inner ear hair cell stereocilia, and to retinal photoreceptors by immunocytochemistry. Our results further strengthen the importance of protocadherin 15 in the morphogenesis and cohesion of stereocilia bundles and retinal photoreceptor cell maintenance or function.

Mutations in a novel gene, TMIE, are associated with hearing loss linked to the DFNB6 locus.

We have identified five different homozygous recessive mutations in a novel gene, TMIE (transmembrane inner ear expressed gene), in affected members of consanguineous families segregating severe-to-profound prelingual deafness, consistent with linkage to DFNB6. The mutations include an insertion, a deletion, and three missense mutations, and they indicate that loss of function of TMIE causes hearing loss in humans. TMIE encodes a protein with 156 amino acids and exhibits no significant nucleotide or deduced amino acid sequence similarity to any other gene.

Mutations of MYO6 are associated with recessive deafness, DFNB37.

Cosegregation of profound, congenital deafness with markers on chromosome 6q13 in three Pakistani families defines a new recessive deafness locus, DFNB37. Haplotype analyses reveal a 6-cM linkage region, flanked by markers D6S1282 and D6S1031, that includes the gene encoding unconventional myosin VI. In families with recessively inherited deafness, DFNB37, our sequence analyses of MYO6 reveal a frameshift mutation (36-37insT), a nonsense mutation (R1166X), and a missense mutation (E216V). These mutations, along with a previously published missense allele linked to autosomal dominant progressive hearing loss (DFNA22), provide an allelic spectrum that probes the relationship between myosin VI dysfunction and the resulting phenotype.