High diagnostic rate of whole genome sequencing in primary ciliary dyskinesia

29 Aim: Primary ciliary dyskinesia (PCD) is a genetic disorder affecting motile cilia. Most cases 30 are inherited recessively, due to variants in more than 50 genes that result in abnormal or 31 absent motile cilia. This leads to chronic upper and lower airway disease, sub-fertility and 32 laterality defects in some cases. Given overlapping clinical features and genetic 33 heterogeneity, diagnosis can be difficult and often occurs late. Of those tested, an estimated 34 30% of genetically screened PCD patients still lack a molecular diagnosis. Here, we aimed to 35 identify how readily a genetic diagnosis could be made in a clinically diagnosed population 36 using whole genome sequencing (WGS) to facilitate identification of pathogenic variants in 37 known genes as well as identify novel PCD candidate genes. 38

WGS was used to screen for variants causing PCD in 8 clinically diagnosed PCD 39 patients, sequenced as trios where parental samples were available. 40

Seven of the eight cases (87.5%) had homozygous or biallelic variants in DNAH5, 41 DNAAF4 or DNAH11 that were classified as pathogenic or likely pathogenic. Three of the 42 variants were deletions, ranging from 3kb to 13kb, for which WGS identified precise 43 breakpoints, permitting confirmation by Sanger sequencing. WGS yielded a high genetic 44 diagnostic rate from this clinically diagnosed population, in part through detection of 45 structural variants as well as identification of a de novo variant in a novel PCD gene TUBB4B. 46

A molecular diagnosis allows for appropriate clinical management for cases and 47 their families, including prediction of phenotypic features correlated to genotype. Here, WGS 48 uplifted genetic diagnosis in cases of clinically diagnosed PCD by identifying structural variants 49 and novel modes of inheritance in new candidate genes. Our study suggests that WGS could 50 be a powerful part of the PCD diagnostic toolkit to increase the current molecular diagnostic 51 yield from 70%. It provides important new insight into our understanding of fundamental 52 biology of motile cilia as well as of variation in the non-coding genome in PCD. 53 54 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 3 55

56 Primary ciliary dyskinesia (PCD, OMIM: PS244400) is a genetic disorder of motile cilia (1–3), 57 highly structurally organised organelles that project from the surface of the cells. Their 58 organised structure allows the cilia to beat in a coordinated manner to effectively propel 59 fluids across the surface of the airways, brain ventricles and reproductive tracts. However, in 60 PCD, these cilia have an abnormal structure or function, resulting in static cilia, cilia that beat 61 in an uncoordinated manner, or a complete absence of cilia (2,3). This can result in a chronic 62 respiratory disease, due to impaired mucociliary clearance, as well as hydrocephaly, laterality 63 defects (e.g. situs inversus) and fertility defects in a subset of patients. PCD often presents at 64 birth as unexplained neonatal respiratory distress. Affected ch ildren then present with 65 variable clinical features including a daily wet cough, chronic respiratory tract infections, 66 rhinitis, sinusitis and otitis media. Over time, these can lead to debilitating long -term 67 complications, including bronchiectasis and hearing impairment (1–6). 68 69 However, PCD is a heterogeneous disorder that has significant phenotypic overlap with other 70 genetic respiratory diseases, such as cystic fibrosis and primary immunodeficiency. The 71 current clinical diagnostic work -up for suspected PCD cases requires a battery of highly 72 specialised tests (7,8). This includes detection of low nasal nitric oxide (nNO), as well as high-73 speed video microscopy (HS VM) to assess cilia beating pattern and frequency and 74 transmission electron microscopy (TEM) to assess the cilia ultrastructure from a nasal brush 75 biopsy. Abnormalities identified in cilia structure and/or beating pattern or in the presence of 76 clinical fea tures such as chronic oto -sinopulmonary symptoms can be used to confirm a 77 diagnosis of PCD (1,2,6,8,9). Access to diagnostic testing is limited to a small number of highly 78 specialised units in the UK. Families must travel for testing, often over long distances and 79 sometimes on multiple occasions. Even in those patients with a clinical diagnosis of PCD, 21% 80 had normal ciliary ultrastructure (10). Therefore, additional sensitive and accessible 81 diagnostic techniques acceptable to patients of all ages are required. 82 83 PCD is a genetically heterogeneous disorder. The majority of cases are autosomal recessively 84 inherited, with over 50 genes currently associated with PCD (1,2,6,11,12). Mutations in these 85 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 4 genes account for approximately 70% of PCD cases screened, s uggesting additional causal 86 genes exist and/or more complicated structural variants (SVs) are being missed in known 87 genes (13). The genes ass ociated with PCD encode proteins involved in cilia structure, cilia 88 assembly or regulatory complexes. There is often a correlation between the affected gene 89 and the defects in cilia motility or ultrastructure observed at diagnosis (1,6,7,9). However, this 90 is not always the case: it is estimated that 9 -20% of PCD cases have normal or in conclusive 91 ultrastructure by TEM (14). Similarly normal or non -diagnostic HSVM results complicate 92 diagnosis for a significant portion of kno wn PCD genes (15). nNO is also considered a useful 93 screening test prior to referral for nasal brushing. While low nNO measurements may be 94 indicative of PCD in patients lacking clear ultrastructural defects, its use, particularly in young 95 children, has not been universally rolled out across PCD centres (16,17) and patients with 96 mutations in at least some PCD genes show normal nNO levels (18,19). In such suspected 97 cases, where a clinical diagnosis is unclear, a genetic diagnosis is an important means to 98 confirm PCD (1,5,6,20). However, genetic testing often occurs late in the diagnostic work -up 99 for PCD (7). Recent guidelines have recommended studies to investigate the utility of genetic 100 testing in the PCD diagnostic pathway, particularl y as next generation sequencing (NGS) is 101 increasingly available (8,21,22). 102 103 In this study, we investi gated the utility of whole genome sequencing (WGS) to detect 104 disease-causing variants in children and young adults with a clinical diagnosis of PCD. We 105 selected a non-endogamous population where PCD was clinically confirmed to test whether 106 WGS could effect ively improve the molecular diagnostic yield of PCD. WGS would provide 107 unbiased genome coverage beyond exons of ~50 genes associated with PCD whilst 108 simultaneously detecting a range of variants, from single nucleotide changes, small indels as 109 well as compl icated SVs, which may be missed by targeted sequencing approaches. Eight 110 patients were recruited, alongside parental samples where available. This led to a genetic 111 diagnosis for all eight cases, three of whom had a pathogenic deletion of 3 kb -13 kb and one 112 patient had a de novo missense variant p.P259L in a novel candidate gene TUBB4B. 113 114 115

116 Patient cohort 117 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 5 Eight patients with a confirmed clinical diagnosis of PCD (50% female), aged 6 to 31 years, 118 were recruited to the study. Signed and in formed consent was obtained from the affected 119 individual as well as relatives through approved protocols. Sample IDs were assigned and the 120 key known only those within the research group. Clinical phenotypes are shown in Table 1. 121 For details of blood samples, DNA extraction, and analysis of sample ethnicity see 122 Supplementary Methods and Supplementary Figure 1 . The study was approved by the 123 London-West London and Gene Therapy Advisory Committee Research Ethics Committee 124 (REC number 11/LO/0883). 125 126 Whole Genome Sequencing 127 DNA was sequenced by WGS at Edinburgh Genomics. Libraries were prep ared using the 128 Illumina TruSeq PCR -free protocol and sequenced on the Illumina HiSeq X platform. The 129 average yield per sample was 136 Gb, with mean coverage of 36x (range 33.9-38.3). 130 131 Gene Panel, data analysis and variant classification 132 A virtual gene pan el of 146 genes was created, which included the known ‘green’ 34 PCD 133 genes plus 107 suspected ciliopathy genes that may have respiratory features based on the 134 PCD PanelApp panel (v1.14) with five additional genes identified in the literature ( CFAP300, 135 DNAH6, DNAJB13, STK36 and TTC25) (23–25). Analysis of variants was carried out as described 136 in Supplementary Methods. Additional analys es were carried out across the FOXJ1 locus for 137 Case 3, in whom no diagnostic variants were identified using the virtual gene panel. A 138 genome-wide expanded analysis identified a de novo missense mutation p.P259L 139 (chr9:g.137242994:C>T (hg38)) in the gene TUBB4B only in the patient, and not present in 140 either parent (26). 141 142 Modelling of whole exome sequencing data 143 A whole exome sequencing (WES) -like subset of the WGS data was obtained by extracting 144 only the reads mapping to the regions in the TWIST Exome Capture Kit (using samtools v1.6) 145 from the BAM file for each sample. WES CNV calling was performed by ExomeDepth (v1.1.15) 146 as detailed in Supplementary Methods. 147 148 Homology modelling of DNAH11 and location of missense variants 149 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 6 A homology model of the C -terminal region of the DNAH11 motor domain (residues 3348 -150 4504) was built using PHYRE2 (27,28). The effects of mutations in the C-terminal domain (CTD) 151 (residues 4124-4504) on protein stability were modelled with FoldX (29). Sequences of human 152 dynein genes were aligned with MUSCLE (30) and the sequence alignment was visualised with 153 MView (31). Further details are given in the Supplementary Methods. 154 155 156

157 Eight PCD patients were recruited to the study as six trios, one proband-mother duo and one 158 singleton (Table 1). All eight cases had a confirmed clinical diagnosis of PCD, following a nasal 159 brush biopsy. In all cases, nasal brushings were undertaken for clinical suspicion of PCD. 160 Presentations and phenotype severity were varied and are summarised in Table 1. 161 162 Using a virtual PCD panel approach to initially screen the WGS, we confirmed genetic 163 diagnosis for seven of the eight cases ( Table 2). Three cases had biallelic pathogenic and/or 164 likely pathogenic variants in the outer dynein arm heavy chain subunit gene DNAH5, which 165 has previously been shown to contribute to the largest proportion of PCD cases among 166 populations of European descent (1). Case 1 had a hemizygous nonsense variant (c.5281C>T, 167 p.(Arg1761Ter)), inherited from the mother and previously reported in PCD cases (32,33), and 168 an overlapping 13kb deletion on the other allele, inherited from the father ( Supplementary 169 Figure 3). Case 2 had two nonsense variants in DNAH5, c.3949C>T, p.(Gln1317Ter), which was 170 previously reported (34), and the c.5281C>T, p.(Arg1761Ter) variant. ddPCR was used to 171 phase the variants in Case 2 (Supplementary Figure 4). Case 4 had a single base pair deletion, 172 resulting in a frameshift and premature termination codon (PTC) (c.10815del, 173 p.(Pro3606Hisfs)), and a single base pair duplication, resulting in direct creation of a PTC 174 (c.13458dup, p.(Asn4487Ter)), which were shown to be in trans through trio -based phasing. 175 Both variants have been reported previously in PCD cases (33). All three cases with biallelic 176 variants in DNAH5 were shown to have absence or defects of outer dynein arms on TEM 177 (Figure 1), with Case 1 and Case 2 also shown to have static cilia ( Table 1, Supplementary 178 Video 1 ). This is consistent with loss of DNAH5, which has been shown to result in outer 179 dynein arm truncation and immotile cilia (33). All three share a similar clinical phenotype, 180 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 7 with chronic wet cough, sinus problems and recurrent chest infections. Hearing loss was also 181 noted in cases 2 and 4. 182 183 Two cases had likely pathogenic variants in the cytoplasmic axonemal dynein assembly factor 184 DNAAF4 (DYX1C1), necessary for assembly and/or stability of the inner and outer dynein arms 185 (Table 2) (35). Case 5 had a homozygous 3.5kb deletion encompassing exon 7 of DNAAF4, 186 which was a lso heterozygous in the mother. There was no evidence of uniparental disomy 187 and, given Case 5 is also reported to have an affected sibling, it is likely the father also carries 188 the 3.5kb deletion, although the father's DNA was unavailable. This variant has previously 189 been reported in PCD (35). Ultrastructural analysis was consistent with previous reports for 190 DNAAF4 mutations, with reports of absent dynein arms (35) (Figure 1 ) and immotile cilia 191 shown on HSVM. Case 6 had a maternally -inherited nonsense va riant (c.856G>T, 192 p.(Glu286Ter)) and a paternally-inherited 3.1 kb deletion encompassing the last two exons of 193 DNAAF4, predicted to disrupt the C -terminal tetratricopeptide -like helical domain region 194 (TPR) by deleting the final 71 amino acids, as well as th e 3’UTR. We predict this allele to be 195 loss-of-function as the resulting transcript is predicted to be subject to nonsense -mediated 196 decay. Case 6 has bronchiectasis with a recurrent need for antibiotics. He has no ear or 197 hearing problems and situs solitus. 198 199 A further two cases had pathogenic or likely pathogenic variants in the outer dynein arm 200 heavy chain subunit DNAH11. Case 7 had a homozygous missense variant in DNAH11 (c.13373 201 C>T, p.(Pro4458Leu)) previously reported in PCD cases (36,37), at a low frequency in gnomAD 202 and in silico predictions support pathogenicity ( Table 2 ). There was sufficient evidence to 203 classify the var iant as likely pathogenic using ACMG guidelines (38,39). Case 8 had a 204 heterozygous 2 bp deletion, resulting in a PTC (c.10 221_10222del, p.(Cys3409Trpfs)), which 205 was classified as pathogenic ( Table 2 ). The second variant was a heterozygous missense 206 (c.13288G>C, p.(Gly4430Arg)). This variant is at a low frequency in gnomAD and in silico 207 predictions support pathogenicity, but it has not previously been reported in PCD cases. As 208 the variant is in trans with the c.10221_10222del variant, shown by phasing with parental 209 samples, there is sufficient evidence to classify the variant as likely pathogenic (Table 2). Both 210 Case 7 and Case 8 were shown to have a normal cilia ultrastructure with a dysmotile 211 phenotype ( Figure 1, Supplementary Video 2 ), consistent with loss of DNAH11, which 212 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 8 localises to the proximal portion of respiratory cilia (40) . There is however phenotypic 213 variability between Case 7 (situs inversus with dextrocardia and recurrent chest infections) 214 and Case 8 (chronic wet cough and recurrent ear infections, with situs solitus). 215 216 The two missense variants in DNAH11 both reside at the 3 ' C-terminal domain (CTD) of the 217 axonemal dynein heavy chain. While there is no structure available for DNAH11, there are 218 homologous structures of other dynein proteins. To investig ate the effects of the DNAH11 219 missense mutations on protein structure, we built a homology model of the DNAH11 motor 220 domain. Both mutations occur within the CTD, on the outer side of the dynein motor dimer 221 (Figure 2A) and are intriguingly very close to each other in three -dimensional space (Figure 222 2B), separated by only 4.2 Å. 223 224 Molecular modelling of the missense mutations using the program FoldX (29) predicts that 225 the Gly4430Arg should be extremely disruptive to protein structure, with a ΔΔG of 10.9 226 kcal/mol. We also modelled all other CTD missense variants presented in the gnomAD v2.1 227 database (41) . Remarkably, out of 269 variants (Supplementary Table 1), Gly4430Arg has the 228 highest ΔΔG and therefore is predicted to be the most damaging. More over, this position is 229 fully conserved across all human dyneins (Figure 2C). This strongly suggests that Gly4430Arg 230 is pathogenic due to its disruptive effects on protein structure. 231 232 In contrast to Gly4430Arg, Pro4458Leu is predicted to be relatively mild at a structural level, 233 with a ΔΔG of 1.1 kcal/mol, making it only the 65 th most damaging out of 269 variants 234 (Supplementary Table 1 ). However, this residue is highly conserved, existing as a proline 235 across all human dyneins except DNAH1, where it is an alanine. Thus, while Pro4458 is unlikely 236 to cause a severe destabilisation of protein structure, its remarkable proximity to Gly4430 237 combined with a moderate structural perturbation and high conservation are supportive of 238 pathogenicity. 239 240 To assess whether the three deletion variants identified in Cases 1, 5 and 6 would have been 241 identified using WES, as opposed to WGS, the WGS data was subsetted to create a WES -like 242 dataset for each sample ( Supplementary Figure 2). CNV calling using ExomeD epth was able 243 to identify the DNAH5 variant c.5272-955_6197del p.(?) in both the proband and father for 244 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 9 Family 1. However, the deletion, which spans five exons and has a breakpoint within exon 37, 245 was called as covering only 4 of the 5 affected exons in the proband. The single exon deletion 246 of DNAAF4 c.784-1037_894-2012del p.(?) was identified for the proband in Family 5, where 247 it is homozygous, but not in the heterozygous mother. The DNAAF4 deletion c.1048 -248 149_*1048del p.(?) identified in Family 6 was not detected in the proband or father, despite 249 both carrying the variant. 250 251 The phenotype of Case 3 is particularly relevant as ultrastructural analysis revealed ciliary 252 agenesis (Supplementary Video 3), sometimes referred to as reduced generation of multiple 253 motile cilia (RGMC), a specific subtype of PCD. In addition, Case 3 has shunted hydrocephalus, 254 having undergone initial ventriculo -peritoneal (VP) shunt insertion neonatally, and a 255 subsequent VP shunt revision as an adolescent. To date, few genes have been implicated in 256 this RGMC phenotype by recessive inheritance, CCNO (cyclin O) and MCIDAS (multicilin) 257 (42,43) but no pathogenic or potentially pathogenic variants were identified in either gene. 258 Heterozygous de novo mutations in the master motile ciliogenesis transcriptional regulator 259 FOXJ1 were identified as the first autosomal dominant cause of a distinct PCD-like condition, 260 associated with chronic respiratory disease, laterality defects and hydrocephalus (18). Similar 261 cellular defects are observed with reduced apical docking of centrioles and fewer cilia, but a 262 focused analysis revealed no identifiable pathogenic mutations in the FOXJ1 locus. An 263 expanded, whole genome analysis for SNP and indel candidates produced a very limited list 264 of variants ( Supplementary Table 2 ). This included a de novo missense mutation p.P259L 265 (chr9:g.137242994:C>T (hg38)) in the gene TUBB4B only in the patient, and not present in 266 either parent or foun d on gnomAD 4.0 or other publicly available databases. Whilst 267 interpretation of pathogenicity from a single patient is limiting, as part of a large international 268 collaboration, we were able to identify a further eleven patients with PCD carrying TUBB4B 269 variants identified by next-generation sequencing (NGS) (26). This included five patients with 270 PCD-only carrying the identical p.P259L (chr9:g. 137242994:C>T) variant, one carrying a 271 different missense p.P259S (chr9:g.137242993:C>T (hg38)) variant and one patient carried an 272 in-frame ten amino acid duplication p.F242_R251dup (chr9: g.137242941_137242970dup 273 (hg38)) (26). Moreover, we also identified a recurrent de novo TUBB4B variant four patients 274 with a p.P358S (chr9:g.137243290:C>T (hg38)) variant, who presented with features of both 275 PCD with Leber congenital amaurosis and sensorineural hearing loss. In -depth functional 276 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 10 analyses in cell and animal models confirmed pathogenicity and distinct dominant -negative 277 mechanisms of action to cause different disease presentations (26). In summary, our WGS 278 strategy was very effective at identifying a novel PCD disease gene with the first report of 279 dominant negative disease mechanisms. 280 281 282 283

284 In this study, WGS of affected probands and family members led to a genetic diagnosis in all 285 eight PCD patients, representing 11 different mutations in 3 autosomal recessive PCD genes 286 and 1 de novo mutation in a novel autosomal dominant PC D candidate TUBB4B. In Case 3, a 287 patient with features of reduced generation of motile cilia and hydrocephalus, no pathogenic 288 variants would been identified in a panel-based approach, even with targeted sequencing of 289 potential candidates for known RGMC loc i. The diagnostic rate of 100% in our small study is 290 higher than previous reports, in which 60-70% of cases received a genetic diagnosis based on 291 the known PCD gene panels (1,5,12,13,20), 75% on extended NGS panels (44) and 68-94% by 292 WES (45–48). 293 294 Next-generation sequencing (NGS) technologies continue to revolutioni se rare genetics 295 research and clinical diagnostics, where the advantages of WES versus WGS are often fiercely 296 debated. Cheaper in terms of costs of sequencing, analysis and data storage, WES is generally 297 preferred as a front -line diagnostic tool. WES, howe ver, has several issues in terms of 298 evenness of genome coverage and sequence bias, particular for copy number variations 299 (CNV). In comparison, several studies have found more accurate variant calls as well as even 300 and unbiased coverage of coding regions ar e generated by WGS (49,50). Our analysis was 301 focused on known and candidate PCD genes, screening simultaneously for single nucleotide 302 variants (SNVs), small indels and more complex SVs. In three cases, pathogenic deletions were 303 identified, ranging in size from 3kb to 13kb. Modelling of our WGS data to represent WES-like 304 data suggested that WES data would only have identified t he deletion in one of these three 305 cases, in which the deletion was homozygous. Since our WES -like model has more uniform 306 coverage than true WES data, our WES-like model could be considered more reliable for CNV 307 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 11 calling than real -life WES sequence. However, we acknowledge that our model is based on 308 36X WGS data, whereas WES would be nearer 100X in practice but coverage does vary across 309 commercial platforms. The ability to detect SVs in PCD is of importance and consistent with 310 recent diagnostic guidelines for PCD, which highlighted that causal SVs and intronic mutations 311 can be missed due to the large number and size of PCD genes (8,21). Further, WES would not 312 have provided the precise breakpoint information that allowed confirmation of the three 313 deletion variants by Sanger sequencing. 314 315 Where WGS was clearly advantageous was PCD disease gene discovery in patients without 316 biallelic variants in known genes or in the one recent example of autosomal dominant 317 inheritance ( FOXJ1) or in the few cases of X -linked recessive inheritance ( RPGR, PIH1D3, 318 OFD1) (12). In Case 3, WES would have likely identified the de novo SNV in TUBB4B, as WES 319 panels identified SNVs in the other 11 TUBB4B patients (26). Here, the clear advantage of 320 WGS was to rule out potential non -coding alterations in the known RGMC PCD genes of 321 known inheritance such as FOXJ1 modes as to quickly prioritise novel candidates in our first 322 proband. As such, a clear benefit of WGS, similar to WES, is that its findings are future-proof; 323 the data generated can be re -screened for PCD -causing variants identified subsequent to 324 genetic testing, if initial testing is inconclusive. However, only WGS will allow fut ure analysis 325 of non-coding genome for variations in regulatory elements such as of transcription factor 326 binding sites that may underlie a subset of unsolved PCD cases. 327 328 Whatever the modality, increased genetic testing for PCD is critical. Currently genetic testing 329 for PCD sits as an additional step to confirm diagnosis by both European and American 330 guidelines (8,21,22). As shown here, and elsewhere, genetic testing can accurately diagnose 331 PCD where standard clinical procedures are unavailable, impractical or inconclusive (45–332 47,51) Importantly, a delayed PCD diagnosis is associated with worse progno sis (52,53). A 333 clinical diagnosis of PCD requires specialised, invasive testing, and frequently necessitates 334 travel over considerable distances to specialist centres. These tests are unsuitable for critically 335 ill neonates and for those unable to travel or unwilling to undergo invasive testing. Genetic 336 testing for PCD may therefore have utility in the neonatal period or in i nfancy that may have 337 been preferred to the diagnostic odyssey of the patients in this study, where the age of 338 diagnosis was between 5 and 12 years. Suggested clinical criteria for genetic testing in PCD 339 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 12 could include term babies who become unwell at >12 ho urs of age with respiratory disease, 340 lobar collapse, situs inversus and/or an unexpectedly high or prolonged oxygen requirement 341 (54,55). Increasing availability, reducing costs and recognised clinical utility of NGS platforms 342 such as WGS in the neonatal and paediatric intensive care unit setting would be consistent 343 with such indications (50,51,54,56–59). Moreover, either WES or WGS would hel p rule out 344 other confounding clinical presentations such as primary immune deficiency disorders (PIDs), 345 with overlapping symptoms of frequent, often severe, airway infections as well as recurrent 346 otitis media, and sinusitis (60–62). Both platforms are advantageous as they allow analysis of 347 all potential causative genes, known and novel, thus faster to keep up with recently reported 348 genes that may not yet be included on PCD-gene panels. Indeed such technologies may prove 349 to be more cost - and time -effective as providing a molecular diagnosis test for PCD than 350 conventional clinical gene panels (45,46,48,63). A prospective study would appear to be 351 warranted, to define the optimum criteria for genetic testing and diagnostic yield in neonates 352 and older infants and children with respiratory symptomatology to help expedite PCD 353 diagnosis. 354 355 While earlier genetic testing for PCD is clearly a priority, PCD also remains underdiagnosed. 356 PCD ha s an estimated incidence of 1 in 7,500 births, rising to 1 in 2,300 in endogamous 357 populations (64,65). In North America, it is estimated that only 1,000 patients have a 358 confirmed diagnosis of PCD as opposed to the predicted ~25,000 - 50,000 these rates would 359 suggest to be affected with PCD (66). Similar underdiagnosis is reported in the UK, where 360 these prevalence estimates suggest there should be at least 8,900 people with PCD in the UK; 361 less than a quarter of these are known to the NHS highly specialised PCD service (67). A large 362 portion of these missing patients are likely adult patients within primary care or 363 bronchiectasis clinics, as suggested by one study which found 12% of bronchiectasis patients 364 had pathogenic variants in known motile ciliopathy genes (68). Increased genetic testing as a 365 first-pass diagnostic test in these suspected cases of PCD could be a way of controlling access 366 to more labour -intensive clinical and pathological diagnostic work -up in limited specialist 367 centres. 368 369 PCD needs to enter the precision medicine era. A genetic diagnosis is key to improved patient 370 prognosis as we better understand genotype-phenotype relations (69) and critically to being 371 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 13 trial-ready as much-needed genetic therapies come online (70,71). Increased genetic testing 372 is also being combined with a curated worldwide database, similar to that of CFTR2 for Cystic 373 Fibrosis (72), to enable a better understanding of these genotype-phenotype relations in PCD 374 called CiliaVar (73). Such a database will significantly facilitate interpretation of variants, of 375 which 21% has been suggested to be variants of unknown significance (VUS) (73). For 376 example, in Cases 7 and 8 in our study, structural modelling of the CTD of DNAH11 aided the 377 assignment of pathogenicity of two missense variants, with one predicted to be detrimental 378 to protein stability and the second, in close proximity to the first, shown to be highly 379 conserved. 380 381 In conclusion, this study demonstrates the benefits of using WGS to obtain a genetic diagnosis 382 for PCD, through its ability to detect SNVs and SVs simultaneously as well as detecting variants 383 in genes outwith current gene panels. The detection of multi -kilobase deletions in three of 384 the seven diagnosed cases highlights the need to detect SVs as part of the genetic testing for 385 PCD. It also allowed rapid prioritisation of SNVs in nov el candidate genes with dominant 386 modes of inheritance. Practically and financially, WGS would likely sit behind current standard 387 of care panel -based or WES diagnostic platforms. Given the high genetic diagnostic rate 388 observed here and elsewhere (44,45,47,48), we suggest that genetic testing should be an 389 early step in the current diagnostic pathway for PCD, particularly in cases where nasal brush 390 biopsy is unavailable. By moving clinical and genetic diagnostic pathways in PCD earlier, 391 ideally to early life, we could have a transformative long -term reduction in morbidity with 392 access to specialist care and disease-modifying therapies commenced before permanent lung 393 damage. 394 395 396

397 We thank the PCD families who participated in this study and the UK PCD Family Support 398 Group for support; Dr Lee Murphy and colleagues at the Wellcome Trust Clinical Research 399 Facility; Edinburgh Genomics for sequencing and analysis; and the Brompton Hospital PCD 400 Diagnostic Service and South East Scotland Genetics Service for clinical support. 401 402 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 14 Support statement: The Scottish Genomes Partnership is funded by the Chief Scientist Office 403 of the Scottish Government Health Directorates [SGP_1] and the MRC Whole Genome 404 Sequencing for Health and Wealth Initiative (MC_PC_15080). We acknowledge support from 405 the MRC (PM: MC_ UU_00007_14, MR_Y015002_1); an MRC Career Development Award 406 (MR_M02122X_1) and Lister Prize Fellowship to JAM; an NHS Research Scotland fellowship 407 to SU; and an NRS/R+D fellowship from the NHS Lothian R&D office to DU. 408 409 Competing interests: We confirm that no competing interests. 410 411 412 413 414 415 416 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 15 Figures: 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 Figure 1: Ultr astructure analysis supports genetic diagnosis for outer arm dynein variants in PCD. 431 (left) DN AH5 variants disrupt outer dynein arms (red arrowhead = disrupted, black arrowheads 432 normal) across axonemes of nasal brush samples from cases HG-001, HG-002 and HG-004. (right) In 433 contrast, DN AH11 variants do not uniformly disrupt outer dynein arms (black arrowheads) in cilia 434 from nasal brush of case HG-007. Disruption of both inner and outer dynein arms (red arrowheads) 435 is observed in DN AAF4 variants, as shown for HG-005. Scale bars = 100 nm. 436 437 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 16 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 Figure 2: Structural and evolutionary analysis of DNAH11 missense mutations. (A ) Structure of the 456 human cytoplasmic dynein-1 dimer (PDB ID: 5NUG), with the location of the C-terminal domain 457 coloured beige, and the equivalent sites of the DNAH11 mutations highlighted in red. (B) Homology 458 model of the DNAH11 C-te rminal domain with the sites of the missense mutations shown in red, 459 along with the ΔΔG values calculated with FoldX. (C) Multiple sequence alignment of human dynein 460 proteins around the region where the missense mutations occur. 461 462 463 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 17 Table 1: Clinical details of PCD cases. 480 Case number Samples recruited Sex Ethnicity Age at diagnosis Age at recruitment Ciliary ultrastructure Cilia motility nNO (ppb) Current FEV1 (% predicted) NRD Chronic wet cough Regular IV antibiotics Recurrent infections Rhinosinusitis Bronchiectasis Recurrent ear infection Hearing loss Situs inversus Dextrocardia 1 Trio F EUR 11- 15 16-20 Absent outer dynein arms Static cilia <5 83 U Y N Y Y N Y N N N 2 Proband only M EUR 6-10 11-15 Outer dynein arm defect Static cilia 8.5 90 Y Y Y N Y Y Y Y N N 3 Trio F EUR 11- 15 11-15 Ciliary agenesis N/A 15.5 47 Y Y Y Y Y Y N Y N N 4 Trio M EUR 0-5 6-10 Outer dynein arm defect Static cilia 75 82 Y Y N Y Y N N Y N N 5 Proband and Mother F EUR 11- 15 31-35 Complete absence of dynein arm Static cilia U U N Y U Y Y Y Y N N N 6 Trio M EUR U 21-25 Unknown Unknown 8.5 62 U Y Y Y Y Y N N N N 7 Trio M SAS 6-10 6-10 Normal Dysmotile 49.5 87 N N N Y Y U N N Y Y 8 Trio F EUR 6-10 11-15 Normal Dysmotile 22 75 U Y N Y Y N Y N N N . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 18 Y= yes; N= no; U= unknown; F= female; M= male; NRD= neonatal respiratory distress; nNO= nasal nitric oxide; ppb= parts per bi llion; FEV1= 481 forced expiratory volume in one second; IV= intravenous; EUR= European; SAS= South Asian 482 483 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 19 Table 2: Genetic variants identified in PCD cases. Variants were classified according to the ACMG/ACGS guidelines (38,39); the criteria met for each variant 468 are shown. 469 Case numbe r Gene Transcript dbSNP ID Variant(s) Zygosity gnomAD allele frequency (v4.0.0) In silico predictions ACMG/ACG S Classificatio n ClinVar Accession Number 1 DNAH5 NM_001369.3 N/A c.5272- 955_6197del p.(?) Compound heterozygous Absenta N/A P (PVS1, PM3, PP4) SCV001334255 rs148891849 c.5281C>T, p.(Arg1761Ter) (32– 34) 6x10-5 N/A P (PVS1, PM2, PM3_str, PP4) SCV001334256 2 DNAH5 NM_001369.3 rs176950857 1 c.3949C>T, p.(Gln1317Ter)(74) Compound heterozygous 1.6x10-6 N/A P (PVS1, PM2, PM3_str, PP4) SCV001334257 rs148891849 c.5281C>T, p.(Arg1761Ter)(32– 34) 6x10-5 N/A P (PVS1, PM2, PM3_str, PP4) SCV001334256 3 TUBB4B NM_006088.6 N/A c.776C>T, p.(Pro259Leu) (26) Heterozygous Absent Grantham Score 98 AlphaMissens e 0.998 (LP) P (PS2, PM2, PP2, PP3) SCV002770069. 1 4 DNAH5 NM_001369.3 rs397515540 c.10815del, p.(Pro3606Hisfs) (33) Compound heterozygous 4.1x10-4 N/A P (PVS1, PM2, PM3_str, PP4) SCV001334258 rs775696136 c.13458dup, p.(Asn4487Ter) (33) 1.6x10-4 N/A P (PVS1, PM2, PM3_str, PP4) SCV001334259 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 20 5 DNAAF4 NM_130810.4 N/A c.784-1037_894- 2012del p.(?) (35) Homozygous 2.5x10-4 b N/A P (PVS1, PM3_str, PP4) SCV001334260 6 DNAAF4 NM_130810.4 rs770136467 c.856G>T, p.(Glu286Ter) Compound heterozygous 2.6x10-5 N/A P (PVS1, PM2, PP4) SCV001334261 N/A c.1048- 149_*1048del p.(?) Absenta N/A P (PVS1, PM3, PP4) SCV001334262 7 DNAH11 NM_00127711 5.2 rs72658835 c.13373C>T, p.(Pro4458Leu) (36,37,47,51) Homozygous 8.4x10-5 Grantham Score 98 REVEL 0.377 (Uncertain) AlphaMissens e 0.625 (LP) LP (PM2, PM3, PP3, PP4) SCV001334263 8 DNAH11 NM_00127711 5.2 rs145054078 8 c.10221_10222del, p.(Cys3409Trpfs) Compound heterozygous 1.5x10-5 N/A P (PVS1, PM2, PP4) SCV001334264 N/A c.13288G>C, p.(Gly4430Arg) (75) 2x10-6 Grantham score 125 REVEL 0.641 (Damaging) AlphaMissens e 0.813 (LP) LP (PM1_sup, PM2, PM3_str, PP3, PP4) SCV001334265 P= pathogenic; LP= likely pathogenic; N/A= not applicable. aabsent from gnomAD SV and gnomAD CNV. bfrequency from gnomAD SV. 470 471 472 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 21 473

474 1. Lucas JS, Davis SD, Omran H, Shoemark A. Primary ciliary dyskinesia in the genomics age. 475 Lancet Respir Med. 2020 Feb;8(2):202–16. 476 2. Leigh MW, Horani A, Kinghorn B, O’Connor MG, Zariwala MA, Knowles MR. Primary Ciliary 477 Dyskinesia (PCD): A genetic disorder of motile cilia. Transl Sci Rare Dis. 2019 Jul 4;4(1–2):51–478 75. 479 3. Wallmeier J, Nielsen KG, Kuehni CE, Lucas JS, Leigh MW, Zariwala MA, et al. Motile 480 ciliopathies. Nat Rev Dis Primers. 2020 Sep 17;6(1):77. 481 4. Damseh N, Quer cia N, Rumman N, Dell SD, Kim RH. Primary ciliary dyskinesia: mechanisms 482 and management. Appl Clin Genet. 2017 Sep 19;10:67–74. 483 5. Mirra V, Werner C, Santamaria F. Primary ciliary dyskinesia: an update on clinical aspects, 484 genetics, diagnosis, and future treatment strategies. Front Pediatr. 2017 Jun 9;5:135. 485 6. Dunsky K, Menezes M, Ferkol TW. Advances in the diagnosis and treatment of primary ciliary 486 dyskinesia: A review. JAMA Otolaryngol Head Neck Surg. 2021 Jun 17; 487 7. Halbeisen FS, Shoemark A, Barbato A, Boon M, Carr S, Crowley S, et al. Time trends in 488 diagnostic testing for primary ciliary dyskinesia in Europe. Eur Respir J. 2019 Oct 24;54(4). 489 8. Shoemark A, Dell S, Shapiro A, Lucas JS. ERS and ATS diagnostic guidelines for primary ciliary 490 dyskinesia: similarities and differences in approach to diagnosis. Eur Respir J. 2019 Sep 491 5;54(3). 492 9. Raidt J, Wallmeier J, Hjeij R, Onnebrink JG, Pennekamp P, Loges NT, et al. Ciliary beat pattern 493 and frequency in genetic variants of primary ciliary dyskinesia. Eur Respir J. 2014 494 Dec;44(6):1579–88. 495 10. Jackson CL, Behan L, Collins SA, Goggin PM, Adam EC, Coles JL, et al. Accuracy of diagnostic496 testing in primary ciliary dyskinesia. Eur Respir J. 2016 Mar;47(3):837–48. 497 11. Horani A, Ferkol TW. Advances in the genetics of primary ciliary dyskinesia: clinical498 implications. Chest. 2018 Sep;154(3):645–52. 499 12. Legendre M, Zaragosi L-E, Mitchison HM. Motile cilia and airway disease. Semin Cell Dev Biol.500 2021 Feb;110:19–33. 501 13. Knowles MR, Zariwala M, Leigh M. Primary Ciliary Dyskinesia. Clin Chest Med. 2016502 Sep;37(3):449–61. 503 14. Goutaki M, Pedersen ESL. Phenotype-genotype associations in primary ciliary dyskinesia:504 where do we stand? Eur Respir J. 2021 Aug 5;58(2). 505 15. Shapiro AJ, Ferkol TW, Manion M, Leigh MW, Davis SD, Knowles MR. High-Speed506 Videomicroscopy Analysis Presents Limitations in Diagnosis of Primary Ciliary Dyskinesia. Am J 507 Respir Crit Care Med. 2020 Jan 1;201(1):122–3. 508 16. Walker WT, Jackson CL, Lackie PM, Hogg C, Lucas JS. Nitric oxide in primary ciliary dyskinesia.509 Eur Respir J. 2012 Oct;40(4):1024–32. 510 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 22 17. Shapiro AJ, Josephson M, Rosenfeld M, Yilmaz O, Davis SD, Polineni D, et al. Accuracy of Nasal 511 Nitric Oxide Measurement as a Diagnostic Test for Primary Ciliary Dyskinesia. A Systematic 512 Review and Meta-analysis. Ann Am Thorac Soc. 2017 Jul;14(7):1184–96. 513 18. Wallmeier J, Frank D, Shoemark A, Nöthe-Menchen T, Cindric S, Olbrich H, et al. De Novo514 Mutations in FOXJ1 Result in a Motile Ciliopathy with Hydrocephalus and Randomization of 515 Left/Right Body Asymmetry. Am J Hum Genet. 2019 Nov 7;105(5):1030–9. 516 19. Raidt J, Krenz H, Tebbe J, Große-Onnebrink J, Olbrich H, Loges NT, et al. Limitations of Nasal517 Nitric Oxide Measurement for Diagnosis of Primary Ciliary Dyskinesia with Normal 518 Ultrastructure. Ann Am Thorac Soc. 2022 Aug;19(8):1275–84. 519 20. Kurkowiak M, Ziętkiewicz E, Witt M. Recent advances in primary ciliary dyskinesia genetics. J520 Med Genet. 2015 Jan;52(1):1–9. 521 21. Lucas JS, Barbato A, Collins SA, Goutaki M, Behan L, Caudri D, et al. European Respiratory522 Society guidelines for the diagnosis of primary ciliary dyskinesia. Eur Respir J. 2017 Jan 4;49(1). 523 22. Shapiro AJ, Davis SD, Polineni D, Manion M, Rosenfeld M, Dell SD, et al. Diagnosis of primary524 ciliary dyskinesia. an official american thoracic society clinical practice guideline. Am J Respir 525 Crit Care Med. 2018 Jun 15;197(12):e24–39. 526 23. Martin AR, Williams E, Foulger RE, Leigh S, Daugherty LC, Niblock O, et al. PanelApp527 crowdsources expert knowledge to establish consensus diagnostic gene panels. Nat Genet. 528 2019 Nov;51(11):1560–5. 529 24. Primary ciliary disorders (Version 1.40) [Internet]. [cited 2024 Feb 17]. Available from:530 https://panelapp.genomicsengland.co.uk/panels/178/. 531 25. Home - OMIM - NCBI [Internet]. [cited 2024 Feb 17]. Available from:532 https://www.ncbi.nlm.nih.gov/omim 533 26. Mechaussier S, Dodd DO, Yeyati PL, McPhie F, Attard T, Shoemark A, et al. TUBB4B variants534 specifically impact ciliary function, causing a ciliopathic spectrum. medRxiv. 2022 Oct 21;20. 535 27. Kelley LA, Mezulis S, Yates CM, Wass MN, Sternberg MJE. The Phyre2 web portal for protein536 modeling, prediction and analysis. Nat Protoc. 2015 Jun;10(6):845–58. 537 28. Zhang K, Foster HE, Rondelet A, Lacey SE, Bahi-Buisson N, Bird AW, et al. Cryo-EM Reveals538 How Human Cytoplasmic Dynein Is Auto-inhibited and Activated. Cell. 2017 Jun 539 15;169(7):1303-1314.e18. 540 29. Guerois R, Nielsen JE, Serrano L. Predicting changes in the stability of proteins and protein541 complexes: a study of more than 1000 mutations. J Mol Biol. 2002 Jul 5;320(2):369–87. 542 30. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput.543 Nucleic Acids Res. 2004 Mar 19;32(5):1792–7. 544 31. Brown NP, Leroy C, Sander C. MView: a web-compatible database search or multiple545 alignment viewer. Bioinformatics. 1998;14(4):380–1. 546 32. Failly M, Bartoloni L, Letourneau A, Munoz A, Falconnet E, Rossier C, et al. Mutations in547 DNAH5 account for only 15% of a non-preselected cohort of patients with primary ciliary 548 dyskinesia. J Med Genet. 2009 Apr;46(4):281–6. 549 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 23 33. Hornef N, Olbrich H, Horvath J, Zariwala MA, Fliegauf M, Loges NT, et al. DNAH5 mutations are 550 a common cause of primary ciliary dyskinesia with outer dynein arm defects. Am J Respir Crit 551 Care Med. 2006 Jul 15;174(2):120–6. 552 34. Hagen EM, Sicko RJ, Kay DM, Rigler SL, Dimopoulos A, Ahmad S, et al. Copy-number variant553 analysis of classic heterotaxy highlights the importance of body patterning pathways. Hum 554 Genet. 2016 Dec;135(12):1355–64. 555 35. Tarkar A, Loges NT, Slagle CE, Francis R, Dougherty GW, Tamayo JV, et al. DYX1C1 is required556 for axonemal dynein assembly and ciliary motility. Nat Genet. 2013 Sep;45(9):995–1003. 557 36. Knowles MR, Leigh MW, Carson JL, Davis SD, Dell SD, Ferkol TW, et al. Mutations of DNAH11558 in patients with primary ciliary dyskinesia with normal ciliary ultrastructure. Thorax. 2012 559 May;67(5):433–41. 560 37. Guan W-J, Li J-C, Liu F, Zhou J, Liu Y-P, Ling C, et al. Next-generation sequencing for identifying561 genetic mutations in adults with bronchiectasis. J Thorac Dis. 2018 May;10(5):2618–30. 562 38. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the563 interpretation of sequence variants: a joint consensus recommendation of the American 564 College of Medical Genetics and Genomics and the Association for Molecular Pathology. 565 Genet Med. 2015 May;17(5):405–24. 566 39. Ellard S. ACGS Best Practice Guidelines for Variant Classification inRare Disease 2020567 [Internet]. 2020 [cited 2024 Feb 17]. Available from: 568 https://www.acgs.uk.com/media/11631/uk-practice-guidelines-for-variant-classification-v4-569 01-2020.pdf570 40. Dougherty GW, Loges NT, Klinkenbusch JA, Olbrich H, Pennekamp P, Menchen T, et al.571 DNAH11 localization in the proximal region of respiratory cilia defines distinct outer dynein 572 arm complexes. Am J Respir Cell Mol Biol. 2016 Aug;55(2):213–24. 573 41. Karczewski KJ, Francioli LC, Tiao G, Cummings BB, Alföldi J, Wang Q, et al. The mutational574 constraint spectrum quantified from variation in 141,456 humans. Nature. 2020 May 575 27;581(7809):434–43. 576 42. Boon M, Wallmeier J, Ma L, Loges NT, Jaspers M, Olbrich H, et al. MCIDAS mutations result in577 a mucociliary clearance disorder with reduced generation of multiple motile cilia. Nat 578 Commun. 2014 Jul 22;5:4418. 579 43. Wallmeier J, Al-Mutairi DA, Chen C-T, Loges NT, Pennekamp P, Menchen T, et al. Mutations in580 CCNO result in congenital mucociliary clearance disorder with reduced generation of multiple 581 motile cilia. Nat Genet. 2014 Jun;46(6):646–51. 582 44. Fassad MR, Patel MP, Shoemark A, Cullup T, Hayward J, Dixon M, et al. Clinical utility of NGS583 diagnosis and disease stratification in a multiethnic primary ciliary dyskinesia cohort. J Med 584 Genet. 2020 May;57(5):322–30. 585 45. Gileles-Hillel A, Mor-Shaked H, Shoseyov D, Reiter J, Tsabari R, Hevroni A, et al. Whole-exome586 sequencing accuracy in the diagnosis of primary ciliary dyskinesia. ERJ Open Research. 2020 587 Oct;6(4). 588 46. Shamseldin HE, Al Mogarri I, Alqwaiee MM, Alharbi AS, Baqais K, AlSaadi M, et al. An exome-589 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 24 first approach to aid in the diagnosis of primary ciliary dyskinesia. Hum Genet. 2020 590 Oct;139(10):1273–83. 591 47. Wheway G, Thomas NS, Carroll M, Coles J, Doherty R, Genomics England Research592 Consortium, et al. Whole genome sequencing in the diagnosis of primary ciliary dyskinesia. 593 BMC Med Genomics. 2021 Sep 23;14(1):234. 594 48. Marshall CR, Scherer SW, Zariwala MA, Lau L, Paton TA, Stockley T, et al. Whole-Exome595 Sequencing and Targeted Copy Number Analysis in Primary Ciliary Dyskinesia. G3 (Bethesda). 596 2015 Jul 2;5(8):1775–81. 597 49. Belkadi A, Bolze A, Itan Y, Cobat A, Vincent QB, Antipenko A, et al. Whole-genome sequencing598 is more powerful than whole-exome sequencing for detecting exome variants. Proc Natl Acad 599 Sci USA. 2015 Apr 28;112(17):5473–8. 600 50. Kingsmore SF, Cakici JA, Clark MM, Gaughran M, Feddock M, Batalov S, et al. A randomized,601 controlled trial of the analytic and diagnostic performance of singleton and trio, rapid genome 602 and exome sequencing in ill infants. Am J Hum Genet. 2019 Oct 3;105(4):719–33. 603 51. Hao C, Guo R, Liu J, Hu X, Guo J, Yao Y, et al. Exome sequencing as the first-tier test for604 pediatric respiratory diseases: A single-center study. Hum Mutat. 2021 Jul;42(7):891–900. 605 52. Halbeisen FS, Pedersen ESL, Goutaki M, Spycher BD, Amirav I, Boon M, et al. Lung function606 from school age to adulthood in primary ciliary dyskinesia. Eur Respir J. 2022 Oct 20;60(4). 607 53. Kuehni CE, Frischer T, Strippoli MPF, Maurer E, Bush A, Nielsen KG, et al. Factors influencing608 age at diagnosis of primary ciliary dyskinesia in European children. Eur Respir J. 2010 609 Dec;36(6):1248–58. 610 54. Nogee LM, Ryan RM. Genetic testing for neonatal respiratory disease. Children (Basel). 2021611 Mar 11;8(3). 612 55. Machogu E, Gaston B. Respiratory Distress in the Newborn with Primary Ciliary Dyskinesia.613 Children (Basel). 2021 Feb 18;8(2). 614 56. French CE, Delon I, Dolling H, Sanchis-Juan A, Shamardina O, Mégy K, et al. Whole genome615 sequencing reveals that genetic conditions are frequent in intensively ill children. Intensive 616 Care Med. 2019 May;45(5):627–36. 617 57. NICUSeq Study Group, Krantz ID, Medne L, Weatherly JM, Wild KT, Biswas S, et al. Effect of618 Whole-Genome Sequencing on the Clinical Management of Acutely Ill Infants With Suspected 619 Genetic Disease: A Randomized Clinical Trial. JAMA Pediatr. 2021 Dec 1;175(12):1218–26. 620 58. Farnaes L, Hildreth A, Sweeney NM, Clark MM, Chowdhury S, Nahas S, et al. Rapid whole-621 genome sequencing decreases infant morbidity and cost of hospitalization. NPJ Genom Med. 622 2018 Apr 4;3:10. 623 59. Petrikin JE, Cakici JA, Clark MM, Willig LK, Sweeney NM, Farrow EG, et al. The NSIGHT1-624 randomized controlled trial: rapid whole-genome sequencing for accelerated etiologic 625 diagnosis in critically ill infants. NPJ Genom Med. 2018 Feb 9;3:6. 626 60. Khalturina EO, Degtyareva ND, Bairashevskaia AV, Mulenkova AV, Degtyareva AV. Modern627 diagnostic capabilities of neonatal screening for primary immunodeficiencies in newborns. 628 Clin Exp Pediatr. 2021 Oct;64(10):504–10. 629 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 25 61. Bousfiha A, Moundir A, Tangye SG, Picard C, Jeddane L, Al-Herz W, et al. The 2022 update of 630 IUIS phenotypical classification for human inborn errors of immunity. J Clin Immunol. 2022 Oct 631 6;42(7):1508–20. 632 62. Hao Y, Hao S, Andersen-Nissen E, Mauck WM, Zheng S, Butler A, et al. Integrated analysis of 633 multimodal single-cell data. Cell. 2021 Jun 24;184(13):3573-3587. 634 63. Platt CD, Zaman F, Bainter W, Stafstrom K, Almutairi A, Reigle M, et al. Efficacy and economics 635 of targeted panel versus whole-exome sequencing in 878 patients with suspected primary 636 immunodeficiency. J Allergy Clin Immunol. 2021 Feb;147(2):723–6. 637 64. O’Callaghan C, Chetcuti P, Moya E. High prevalence of primary ciliary dyskinesia in a British 638 Asian population. Arch Dis Child. 2010 Jan;95(1):51–2. 639 65. Hannah WB, Seifert BA, Truty R, Zariwala MA, Ameel K, Zhao Y, et al. The global prevalence 640 and ethnic heterogeneity of primary ciliary dyskinesia gene variants: a genetic database 641 analysis. Lancet Respir Med. 2022 May;10(5):459–68. 642 66. O’Connor MG, Mosquera R, Metjian H, Marmor M. CHEST Reviews: Primary Ciliary Dyskinesia. 643 CHEST Pulmonary. 2023; 644 67. NHS England Highly Specialised Services 2020/21 [Internet]. [cited 2024 Feb 17]. Available 645 from: https://www.england.nhs.uk/wp-content/uploads/2023/11/PRN00328-highly-646 specialised-services-2020-2021.pdf 647 68. Shoemark A, Griffin H, Wheway G, Hogg C, Lucas JS, Genomics England Research Consortium, 648 et al. Genome sequencing reveals underdiagnosis of primary ciliary dyskinesia in 649 bronchiectasis. Eur Respir J. 2022 Nov 17;60(5). 650 69. Shoemark A, Rubbo B, Legendre M, Fassad MR, Haarman EG, Best S, et al. Topological data 651 analysis reveals genotype-phenotype relationships in primary ciliary dyskinesia. Eur Respir J. 652 2021 Aug 5;58(2). 653 70. Raidt J, Maitre B, Pennekamp P, Altenburg J, Anagnostopoulou P, Armengot M, et al. The 654 disease-specific clinical trial network for primary ciliary dyskinesia: PCD-CTN. ERJ Open 655 Research. 2022 Jul;8(3). 656 71. ReCode doses first subject in Phase I primary ciliary dyskinesia trial [Internet]. [cited 2024 Feb 657 17]. Available from: https://www.clinicaltrialsarena.com/news/recode-primary-ciliary-658 dyskinesia-trial/ 659 72. Welcome to CFTR2 | CFTR2 [Internet]. [cited 2024 Feb 17]. Available from: https://cftr2.org/ 660 73. Mani R, Gomes M, González AR, Hogg C, Morris-Rosendahl D, Maitre B, et al. Development 661 and first results of the BEAT-PCD international Primary Ciliary Dyskinesia gene variant 662 database: CiliaVar. Paediatric respiratory epidemiology. European Respiratory Society; 2021. 663 p. PA3458. 664 74. Aprea I, Nöthe-Menchen T, Dougherty GW, Raidt J, Loges NT, Kaiser T, et al. Motility of 665 efferent duct cilia aids passage of sperm cells through the male reproductive system. Mol 666 Hum Reprod. 2021 Feb 27;27(3). 667 75. Leung GKC, Mak CCY, Fung JLF, Wong WHS, Tsang MHY, Yu MHC, et al. Identifying the genetic 668 causes for prenatally diagnosed structural congenital anomalies (SCAs) by whole-exome 669 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 26 sequencing (WES). BMC Med Genomics. 2018 Oct 25;11(1):93. 670 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint High diagnostic rate of whole genome sequencing in primary ciliary dyskinesia Supplementary Material and Methods Holly A Black 1,2, Sophie Marion de Proce 1*, Jose L Campos 3*, Alison Meynert 3, Mihail Halachev 3, Joseph A Marsh 3, Robert A Hirst 4, Chris O’Callaghan 4, Scottish Genomes Partnership, Javier Santoyo-Lopez5, Jennie Murray 2,3, Kenneth Macleod 6, Don S Urquhart 6,7, Stefan Unger 6,7, Timothy J Aitman1^, Pleasantine Mill3^ 1 Centre for Genomic and Experimental Medicine, MRC Institute of Genetics and Cancer, University of Edinburgh, Edinburgh, UK 2 South East of Scotland Genetics Service, Western General Hospital, Edinburgh, UK 3 MRC Human Genetics Unit, MRC Institute of Genetics and Cancer, University of Edinburgh, Edinburgh, UK 4 Centre for PCD Diagnosis and Research, Department of Respiratory Sciences, University of Leicester, UK 5 Edinburgh Genomics, Edinburgh, UK 6 Department of Paediatric Respiratory and Sleep Medicine, Royal Hospital for Sick Children, Edinburgh, UK 7 Department of Child Life and Health, University of Edinburgh, Edinburgh, UK * These authors contributed equally to the manuscript ^ Joint senior authors 1 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint SUPPLEMENTARY MATERIALS Supplementary Table 1: Ranked list of most disruptive reported variants ( ΔΔG) in the C-terminal domain (CTD) of DNAH11 as predicted by FoldX. ΔΔG represents the change in free energy by mutation/design of proteins as predicted by the FoldX algorithm, where ΔΔG = ΔGfold(mutation) − ΔGfold(wild type). Supplementary Table 2: Genome wide list of variants detected in patient HG-003 by Slivar. Supplementary File 1: Gene panel used for variant filtering with G2P under a biallelic inheritance model Supplementary File 2: Gene panel used for variant filtering with G2P under a monoallelic inheritance model Supplementary File 3 Droplet digital PCR for variant phasing. Supplementary Video 1: HSVM for cases 1, 2 and 4, which have a genetic diagnosis in DNAH5, showing static cilia Supplementary Video 2: HSVM for cases 7 and 8, which have a genetic diagnosis in DNAH11, showing dysmotile cilia Supplementary Video 3: HSVM for case 3, showing ciliary agenesis 2 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint Supplementary Figure 1: Principal component analysis (PCA) of study samples compared to 1000 Genomes project samples. Peddy was used to predict ancestry of the samples used in the study by comparison with the 1000 Genomes samples. The 1000 Genomes samples (dots) are colour-coded by location. The samples in our study are represented by squares. All samples were predicted to be of European ancestry (purple squares), except three (orange squares), which were predicted to have South Asian ancestry. AFR= African; AMR= American; EAS= East Asian; EUR= European; SAS= South Asian. 3 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint Supplementary Figure 2: Alignments of the modelled WES data showing sufficient coverage to call the deletion variants identified in Cases 1, 5 and 6. A: Alignment for Family 1, showing coverage of exons 32-38 of DNAH5. B: Alignment for Family 5, showing coverage of exons 6 to 8 of DNAAF4. C: Alignment for Family 6, showing coverage exons 8-10 ofDNAAF4. 4 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint Supplementary Figure 3: Sanger sequencing confirms 13kb deletion in DNAH5 in Case 1. Alignments of the WGS data for Family 1 show a drop in read depth to approximately 50% of that of the surrounding regions across a 13kb region of DNAH5 in the proband and the father. This spans from intron 32 to exon 37. PCR and Sanger sequencing across the breakpoints confirms this deletion. * indicates the position of the c.5281C>T nonsense variant, which is on the maternal haplotype and is therefore hemizygous in the proband. 5 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint Supplementary Figure 4: Drop phase results confirms the two DNAH5 nonsense variants are on different haplotypes for Case 2 . The c.3949 variant was assayed using FAM probes and the c.5281 variant was assayed using HEX probes. For each combination of alleles, a representative result from genomic DNA (left) and PacI-digested DNA (right) is shown. Each figure plots the number of FAM-only positive (blue), HEX-only positive (green), FAM and HEX-positive (orange) and negative (grey) droplets. The linkage % is shown for each test. (Details see Supplementary File 3). 6 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint SUPPLEMENTARY METHODS Patient cohort Eight young people with PCD (50% female), aged 6 to 31 years (mean=15, SD= 7.9), were recruited to the study within the Department of Paediatric Respiratory and Sleep Medicine at the Royal Hospital for Sick Children, Edinburgh, and the South East Scotland Genetics service. All eight cases had a confirmed clinical diagnosis of PCD, following TEM and/or HSVM of a nasal brush biopsy sample. The screening and diagnostic testing was performed according to the PCD National Service protocols, with investigations including nNO, nasal brush biopsies analysed by HSVM for ciliary beat frequency and pattern and quantitative electron microscopy for ciliary ultrastructure. Clinical phenotypes are shown in Table 1. Blood samples were collected in EDTA tubes from the patients and parents, where available. DNA was extracted using the Chemagic DNA blood kit (Chemagen) or the Nucleon Bacc3 kit (GE Healthcare). Sample ethnicity was assessed using Peddy (v4.0.6) (1) (Supplementary Figure 1). Gene Panel, sequence data analysis and variant classification BCBio-Nextgen (0.9.7) was used for alignment and variant detection. This used bwa mem (0.7.13) to align reads to the hg38/GRCh38 reference genome (2), samblaster (0.1.22) to mark duplicate fragments (3) and GATK (3.4-0-g7e26428) for indel realignment and base recalibration (4). GATK HaplotypeCaller was used to calculate genotype likelihoods. Joint genotyping and quality control, including kinship estimates to confirm sample relatedness, were performed using in-house pipelines with GATK (4.0.2.1) following the GATK best practices. Variants were annotated using Ensembl variant effect predictor (VEP 90) (5). A bespoke gene panel of 146 genes was created (Supplementary Files 1 and 2), based on the PCD PanelApp panel (v1.14) and five additional genes identified in the literature (CFAP300, DNAH6, DNAJB13, STK36 and TTC25) (6–8). Variants (SNPs and small indels) were filtered to retain only those within genes on the gene panel and then further filtered to identify candidate variants by inheritance model (both biallelic and monoallelic), transcript consequence, and population allele frequency using the G2P plugin for VEP (9). Variants were assessed using Alamut (v2.13) (10) and classified using the ACMG variant 7 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint interpretation guidelines (11,12). Any variants classified as pathogenic or likely pathogenic were validated using Sanger sequencing and were submitted to ClinVar. Parental samples were used to determine the phase of compound heterozygous variants, except for Case 2, which used droplet digital PCR (ddPCR) for phasing (Supplementary File 3). SVs were called using Manta (13) and Canvas (version 1.38) (14). The detection of SVs can be challenging and it is common practise to use complementary approaches to detect them (15). Manta detects SVs using discordant paired-end and split reads, whereas Canvas relies on changes in read coverage. We searched for any SVs present in our gene panel that were inherited from either parent. We also searched for de novo SVs for cases where both parents were available. SVs were confirmed in silico using SV-Plaudit (16), a tool for rapidly curating SV predictions, and/or using the Integrative Genomics Viewer tool (17). Candidate variants were confirmed in the laboratory by PCR and Sanger sequencing across the deletion breakpoints. For Case 3, we did not find any diagnostic variants using our gene panel. As FOXJ1 was only recently identified as a PCD gene and hence was not present on the panel, we searched for SNVs, indels and SVs in Case 3 in this gene, given the ciliary agenesis phenotype observed in this case is associated with FOXJ1. We also expanded our analysis for this case to a genome wide search for SNV and small indel candidates with Slivar (0.1.10), following the protocol for rare diseases (https://github.com/brentp/slivar/wiki/rare-disease Date accessed: January 2020). A small number of variants was detected (Supplementary Table 2) and none were identified that fitted the current modes of inheritance for PCD. Modelling of whole exome sequencing data A whole exome sequencing (WES)-like subset of the WGS data was obtained by extracting only the reads mapping to the regions in the TWIST Exome Capture Kit (using samtools v1.6) from the BAM file for each sample. The capture region fully covers the exons affected by copy number variants (CNVs) identified based on the WGS data (Supplementary Figure 2). The WES CNV calling was performed using ExomeDepth (v 1.1.15) separately on each individual from the three families in which a pathogenic CNV was identified (Families 1, 5 8 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint and 6). As controls, we used WES-like subset data from the WGS data for the other samples in this project (total of 21), from which we excluded any members of the family currently being evaluated. As the WGS data was aligned to GRCh38, we generated a custom reference dataset (exons.GRCh38) required by ExomeDepth, to replace the dataset currently distributed with the ExomeDepth package (exons.hg19); exons.GRCh38 is based on the latest CCDS release (r 22) available for the GRCh38 human genome reference (Available at: www.ncbi.nlm.nih.gov/projects/CCDS/CcdsBrowse.cgi?REQUEST=SHOW_STATISTICS). Homology modelling of DNAH11 and location of missense variants A homology model of the C-terminal region of the DNAH11 motor domain (residues 3348-4504) was built using PHYRE2 (18), based upon the cryo-electron microscopy structure of human cytoplasmic dynein-1 (PDB ID: 5NUG) (19). The effects of the mutations in the C-terminal domain (CTD) (residues 4124-4504) on protein stability were modelled with FoldX (20), using default parameters and calculated over 10 replicates. Sequences of human dynein genes were aligned with MUSCLE (21) and the sequence alignment was visualised with MView (22). 9 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint SUPPLEMENTARY REFERENCES 1. Pedersen BS, Quinlan AR. Who’s Who? Detecting and Resolving Sample Anomalies in Human DNA Sequencing Studies with Peddy. Am J Hum Genet. 2017 Mar 2;100(3):406–13. 2. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009 Jul 15;25(14):1754–60. 3. Faust GG, Hall IM. SAMBLASTER: fast duplicate marking and structural variant read extraction. Bioinformatics. 2014 Sep 1;30(17):2503–5. 4. McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, et al. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010 Sep;20(9):1297–303. 5. McLaren W, Gil L, Hunt SE, Riat HS, Ritchie GRS, Thormann A, et al. The ensembl variant effect predictor. Genome Biol. 2016 Jun 6;17(1):122. 6. Martin AR, Williams E, Foulger RE, Leigh S, Daugherty LC, Niblock O, et al. PanelApp crowdsources expert knowledge to establish consensus diagnostic gene panels. Nat Genet. 2019 Nov;51(11):1560–5. 7. Home - OMIM - NCBI [Internet]. [cited 2024 Feb 17]. Available from: https://www.ncbi.nlm.nih.gov/omim 8. Primary ciliary disorders (Version 1.40) [Internet]. [cited 2024 Feb 17]. Available from: https://panelapp.genomicsengland.co.uk/panels/178/ 9. Thormann A, Halachev M, McLaren W, Moore DJ, Svinti V, Campbell A, et al. Flexible and scalable diagnostic filtering of genomic variants using G2P with Ensembl VEP . Nat Commun. 2019 May 30;10(1):2373. 10. Alamut TM Visual Plus - Variant Annotation and Analysis Software [Internet]. [cited 2024 Feb 18]. Available from: https://www.sophiagenetics.com/platform/alamut-visual-plus/ 11. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015 May;17(5):405–24. 12. Ellard S. ACGS Best Practice Guidelines for Variant Classification inRare Disease 2020 [Internet]. 2020 [cited 2024 Feb 17]. Available from: https://www.acgs.uk.com/media/11631/uk-practice-guidelines-for-variant-classification-v4-01 -2020.pdf 13. Chen X, Schulz-Trieglaff O, Shaw R, Barnes B, Schlesinger F, Källberg M, et al. Manta: rapid detection of structural variants and indels for germline and cancer sequencing applications. Bioinformatics. 2016 Apr 15;32(8):1220–2. 14. Ivakhno S, Roller E, Colombo C, Tedder P , Cox AJ. Canvas SPW: calling de novo copy number variants in pedigrees. Bioinformatics. 2018 Feb 1;34(3):516–8. 15. Cameron DL, Di Stefano L, Papenfuss AT. Comprehensive evaluation and characterisation of short read general-purpose structural variant calling software. Nat Commun. 2019 Jul 10 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint 19;10(1):3240. 16. Belyeu JR, Nicholas TJ, Pedersen BS, Sasani TA, Havrilla JM, Kravitz SN, et al. SV-plaudit: A cloud-based framework for manually curating thousands of structural variants. Gigascience. 2018 Jul 1;7(7). 17. Robinson JT, Thorvaldsdóttir H, Winckler W, Guttman M, Lander ES, Getz G, et al. Integrative genomics viewer. Nat Biotechnol. 2011 Jan;29(1):24–6. 18. Kelley LA, Mezulis S, Yates CM, Wass MN, Sternberg MJE. The Phyre2 web portal for protein modeling, prediction and analysis. Nat Protoc. 2015 Jun;10(6):845–58. 19. Zhang K, Foster HE, Rondelet A, Lacey SE, Bahi-Buisson N, Bird AW, et al. Cryo-EM Reveals How Human Cytoplasmic Dynein Is Auto-inhibited and Activated. Cell. 2017 Jun 15;169(7):1303-1314.e18. 20. Guerois R, Nielsen JE, Serrano L. Predicting changes in the stability of proteins and protein complexes: a study of more than 1000 mutations. J Mol Biol. 2002 Jul 5;320(2):369–87. 21. Edgar RC. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 2004 Mar 19;32(5):1792–7. 22. Brown NP , Leroy C, Sander C. MView: a web-compatible database search or multiple alignment viewer. Bioinformatics. 1998;14(4):380–1. 11 . CC-BY 4.0 International licenseIt is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint The copyright holder for thisthis version posted February 22, 2024. ; https://doi.org/10.1101/2024.02.21.24302995doi: medRxiv preprint