Rare variant association analyses reveal the significant contribution of carbohydrate metabolic disturbance in severe adolescent idiopathic scoliosis.

BACKGROUND: Adolescent idiopathic scoliosis (AIS), the predominant genetic-influenced scoliosis, results in spinal deformities without vertebral malformations. However, the molecular aetiology of AIS remains unclear. METHODS: Using genome/exome sequencing, we studied 368 patients with severe AIS (Cobb angle >40°) and 3794 controls from a Han Chinese cohort. We performed gene-based and pathway-based weighted rare variant association tests to assess the mutational burden of genes and established biological pathways. Differential expression analysis of muscle tissues from 14 patients with AIS and 15 controls was served for validation. RESULTS: SLC16A8, a lactate transporter linked to retinal glucose metabolism, was identified as a novel severe AIS-associated gene (p=3.08E-06, false discovery rate=0.009). Most AIS cases with deleterious SLC16A8 variants demonstrated early onset high myopia preceding scoliosis. Pathway-based burden test also revealed a significant enrichment in multiple carbohydrate metabolism pathways, especially galactose metabolism. Patients with deleterious variants in these genes demonstrated a significantly larger spinal curve. Genes related to catabolic processes and nutrient response showed divergent expression between AIS cases and controls, reinforcing our genomic findings. CONCLUSION: This study uncovers the pivotal role of genetic variants in carbohydrate metabolism in the development of AIS, unveiling new insights into its aetiology and potential treatment.

Core planar cell polarity genes VANGL1 and VANGL2 in predisposition to congenital vertebral malformations.

Congenital scoliosis (CS), affecting approximately 0.5 to 1 in 1,000 live births, is commonly caused by congenital vertebral malformations (CVMs) arising from aberrant somitogenesis or somite differentiation. While Wnt/ß-catenin signaling has been implicated in somite development, the function of Wnt/planar cell polarity (Wnt/PCP) signaling in this process remains unclear. Here, we investigated the role of Vangl1 and Vangl2 in vertebral development and found that their deletion causes vertebral anomalies resembling human CVMs. Analysis of exome sequencing data from multiethnic CS patients revealed a number of rare and deleterious variants in VANGL1 and VANGL2, many of which exhibited loss-of-function and dominant-negative effects. Zebrafish models confirmed the pathogenicity of these variants. Furthermore, we found that Vangl1 knock-in (p.R258H) mice exhibited vertebral malformations in a Vangl gene dose- and environment-dependent manner. Our findings highlight critical roles for PCP signaling in vertebral development and predisposition to CVMs in CS patients, providing insights into the molecular mechanisms underlying this disorder.

Unraveling the genetic architecture of congenital vertebral malformation with reference to the developing spine.

Congenital vertebral malformation, affecting 0.13-0.50 per 1000 live births, has an immense locus heterogeneity and complex genetic architecture. In this study, we analyze exome/genome sequencing data from 873 probands with congenital vertebral malformation and 3794 control individuals. Clinical interpretation identifies Mendelian etiologies in 12.0% of the probands and reveals a muscle-related disease mechanism. Gene-based burden test of ultra-rare variants identifies risk genes with large effect sizes (ITPR2, TBX6, TPO, H6PD, and SEC24B). To further investigate the biological relevance of the genetic association signals, we perform single-nucleus RNAseq on human embryonic spines. The burden test signals are enriched in the notochord at early developmental stages and myoblast/myocytes at late stages, highlighting their critical roles in the developing spine. Our work provides insights into the developmental biology of the human spine and the pathogenesis of spine malformation.

SIGMA leverages protein structural information to predict the pathogenicity of missense variants.

Leveraging protein structural information to evaluate pathogenicity has been hindered by the scarcity of experimentally determined 3D protein. With the aid of AlphaFold2 predictions, we developed the structure-informed genetic missense mutation assessor (SIGMA) to predict missense variant pathogenicity. In comparison with existing predictors across labeled variant datasets and experimental datasets, SIGMA demonstrates superior performance in predicting missense variant pathogenicity (AUC = 0.933). We found that the relative solvent accessibility of the mutated residue contributed greatly to the predictive ability of SIGMA. We further explored combining SIGMA with other top-tier predictors to create SIGMA+, proving highly effective for variant pathogenicity prediction (AUC = 0.966). To facilitate the application of SIGMA, we pre-computed SIGMA scores for over 48 million possible missense variants across 3,454 disease-associated genes and developed an interactive online platform (https://www.sigma-pred.org/). Overall, by leveraging protein structure information, SIGMA offers an accurate structure-based approach to evaluating the pathogenicity of missense variants.

Impaired glycine neurotransmission causes adolescent idiopathic scoliosis.

Adolescent idiopathic scoliosis (AIS) is the most common form of spinal deformity, affecting millions of adolescents worldwide, but it lacks a defined theory of etiopathogenesis. Because of this, treatment of AIS is limited to bracing and/or invasive surgery after onset. Preonset diagnosis or preventive treatment remains unavailable. Here, we performed a genetic analysis of a large multicenter AIS cohort and identified disease-causing and predisposing variants of SLC6A9 in multigeneration families, trios, and sporadic patients. Variants of SLC6A9, which encodes glycine transporter 1 (GLYT1), reduced glycine-uptake activity in cells, leading to increased extracellular glycine levels and aberrant glycinergic neurotransmission. Slc6a9 mutant zebrafish exhibited discoordination of spinal neural activities and pronounced lateral spinal curvature, a phenotype resembling human patients. The penetrance and severity of curvature were sensitive to the dosage of functional glyt1. Administration of a glycine receptor antagonist or a clinically used glycine neutralizer (sodium benzoate) partially rescued the phenotype. Our results indicate a neuropathic origin for "idiopathic" scoliosis, involving the dysfunction of synaptic neurotransmission and central pattern generators (CPGs), potentially a common cause of AIS. Our work further suggests avenues for early diagnosis and intervention of AIS in preadolescents.

Exploring the association between congenital vertebral malformations and neural tube defects.

Congenital vertebral malformations (CVMs) and neural tube defects (NTDs) are common birth defects affecting the spine and nervous system, respectively, due to defects in somitogenesis and neurulation. Somitogenesis and neurulation rely on factors secreted from neighbouring tissues and the integrity of the axial structure. Crucial signalling pathways like Wnt, Notch and planar cell polarity regulate somitogenesis and neurulation with significant crosstalk. While previous studies suggest an association between CVMs and NTDs, the exact mechanism underlying this relationship remains unclear. In this review, we explore embryonic development, signalling pathways and clinical phenotypes involved in the association between CVMs and NTDs. Moreover, we provide a summary of syndromes that exhibit occurrences of both CVMs and NTDs. We aim to provide insights into the potential mechanisms underlying the association between CVMs and NTDs, thereby facilitating clinical diagnosis and management of these anomalies.

Retrospective Analysis of Associated Anomalies in 636 Patients with Operatively Treated Congenital Scoliosis.

BACKGROUND: Congenital scoliosis is frequently associated with anomalies in multiple organ systems. However, the prevalence and distribution of associated anomalies remain unclear, and there is a large amount of variation in data among different studies. METHODS: Six hundred and thirty-six Chinese patients who had undergone scoliosis correction surgery at Peking Union Medical College Hospital from January 2012 to July 2019 were recruited, as a part of the Deciphering disorders Involving Scoliosis and COmorbidities (DISCO) study. The medical data for each subject were collected and analyzed. RESULTS: The mean age (and standard deviation) at the time of presentation for scoliosis was 6.4 ± 6.3 years, and the mean Cobb angle of the major curve was 60.8° ± 26.5°. Intraspinal abnormalities were found in 186 (30.3%) of 614 patients, with diastematomyelia being the most common anomaly (59.1%; 110 of 186). The prevalence of intraspinal abnormalities was remarkably higher in patients with failure of segmentation and mixed deformities than in patients with failure of formation (p < 0.001). Patients with intraspinal anomalies showed more severe deformities, including larger Cobb angles of the major curve (p < 0.001). We also demonstrated that cardiac anomalies were associated with remarkably worse pulmonary function, i.e., lower forced expiratory volume in the first second (FEV1), forced vital capacity (FVC), and peak expiratory flow (PEF). Additionally, we identified associations among different concomitant malformations. We found that patients with musculoskeletal anomalies of types other than intraspinal and maxillofacial were 9.2 times more likely to have additional maxillofacial anomalies. CONCLUSIONS: In our cohort, comorbidities associated with congenital scoliosis occurred at a rate of 55%. To our knowledge, our study is the first to show that patients with congenital scoliosis and cardiac anomalies have reduced pulmonary function, as demonstrated by lower FEV1, FVC, and PEF. Moreover, the potential associations among concomitant anomalies revealed the importance of a comprehensive preoperative evaluation scheme. LEVEL OF EVIDENCE: Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.

Artificially Intelligent Three-Dimensionally-Printed Patient-Specific Instrument Improves Total Hip Arthroplasty Accuracy.

BACKGROUND: Patient-specific instrumentation (PSI) has the potential to improve the accuracy of implant positioning in total hip arthroplasty (THA). This prospective clinical study aimed to develop artificial intelligence to increase PSI production efficiency and assess accuracy, clinical outcomes, and learning curves. METHODS: A convolutional neural network was applied to automatically process computer tomography images. PSI size and position were designed to guide the acetabular preparation and femoral neck resection. Thirty patients who underwent PSI-assisted THAs were matched to thirty patients who underwent free-hand THAs, and the component positions, as well as radiographic and clinical outcomes were analyzed. RESULTS: PSI-assisted THA was significantly more accurate than free-hand THA at achieving the target component position. The mean absolute errors of cup inclination (P = .004) and anteversion (P < .001) were significantly smaller in the PSI group with fewer outliers. Calcar length (P = .002) and neck length (P = .026) were also more accurate in the PSI group. The leg length discrepancy was significantly lower in the PSI group (P = .002). There were no significant differences in operation time, blood loss, leg length discrepancy, or cup position among the first, second, and last 10 cases. CONCLUSION: PSI-assisted THA offered more accurate component positions and better radiographic outcomes than free-hand THA. There was no evidence of a learning curve. Our findings suggest that PSI is a convenient and practical option to help surgeons achieve accurate surgical outcomes.

Functional characteristics of a broad spectrum of TBX6 variants in Mayer-Rokitansky-Küster-Hauser syndrome.

PURPOSE: Mayer-Rokitansky-Küster-Hauser syndrome (MRKHS) is characterized by congenital absence of the uterus, cervix, and the upper part of the vagina in females. Whole-gene deletion and loss-of-function variants in TBX6 have been identified in association with MRKHS. We aimed to expand the spectrum of TBX6 variants in MRKHS and explore the biological effect of the variant alleles. METHODS: Rare variants in TBX6 were called from a combined multiethnic cohort of 622 probands with MRKHS who underwent exome sequencing or genome sequencing. Multiple in vitro functional experiments were performed, including messenger RNA analysis, western blotting, transcriptional activity assay, and immunofluorescence staining. RESULTS: We identified 16 rare variants in TBX6 from the combined cohort, including 1 protein-truncating variant reported in our previous study and 15 variants with unknown effects. By comparing the prevalence of TBX6 variants in the Chinese MRKHS cohort vs 1038 female controls, we observed a significant mutational burden of TBX6 in affected individuals (P = .0004, odds ratio = 5.25), suggesting a causal role of TBX6 variants in MRKHS. Of the 15 variants with uncertain effects, 7 were shown to induce a loss-of-function effect through various mechanisms. The c.423G>A (p.Leu141=) and c.839+5G>A variants impaired the normal splicing of TBX6 messenger RNA, c.422T>C (p.Leu141Pro) and c.745G>A (p.Val249Met) led to decreased protein expression, c.10C>T (p.Pro4Ser) and c.400G>A (p.Glu134Lys) resulted in perturbed transcriptional activity, and c.356G>A (p.Arg119His) caused protein mislocalization. We observed incomplete penetrance and variable expressivity in families carrying deleterious variants, which indicates a more complex genetic mechanism than classical Mendelian inheritance. CONCLUSION: Our study expands the mutational spectrum of TBX6 in MRKHS and delineates the molecular pathogenesis of TBX6 variants, supporting the association between deleterious variants in TBX6 and MRKHS.

Does Abnormal Preoperative Coagulation Status Lead to More Perioperative Blood Loss in Spinal Deformity Correction Surgery?

This study aims to analyze the potential association between the preoperative coagulation status and perioperative blood loss in spinal deformity correction surgery. The preoperative coagulation status and estimated blood loss (EBL) during operation, postoperative wound drainage, and allogeneic transfusion during and after operation were recorded and analyzed. Among the 164 patients, 26 had a longer prothrombin time (PT), 13 had a lower fibrinogen level, 55 had a longer activated partial thromboplastin time (APTT), and 2 had a longer thrombin time (TT), and the platelet count (PLT) was all normal or higher than the normal level. The mean EBL per surgical level was 77.8 ml (range, 22-267 ml), and the mean drainage per surgical level was 52.7 ml (range, 7-168 ml). Fifty-five patients and 12 patients underwent allogeneic transfusion during and after the operation, respectively. The differences in EBL per surgical level, mean drainage per surgical level, the occurrences of allogeneic transfusion during and after operation between the patients with a longer PT, lower fibrinogen level, longer APTT or longer TT, and the normal controls were not significant (all P's > 0.05). The Spearman correlation analysis showed that there was no correlation between PT, fibrinogen, APTT, TT or PLT with EBL per surgical level, mean drainage per surgical level, or allogeneic transfusion during and after the operation (all P's > 0.05). The abnormal preoperative coagulation status but not hemophilia does not lead to more perioperative blood loss or a higher rate of perioperative allogeneic transfusion in spinal deformity correction surgery.

Exosomes derived from magnetically actuated bone mesenchymal stem cells promote tendon-bone healing through the miR-21-5p/SMAD7 pathway.

Graft healing after anterior cruciate ligament reconstruction (ACLR) involves slow biological processes, and various types of biological modulations have been explored to promote tendon-to-bone integration. Exosomes have been extensively studied as a promising new cell-free strategy for tissue regeneration, but few studies have reported their potential in tendon-to-bone healing. In this study, a novel type of exosome derived from magnetically actuated (iron oxide nanoparticles (IONPs) combined with a magnetic field) bone mesenchymal stem cells (BMSCs) (IONP-Exos) was developed, and the primary purpose of this study was to determine whether IONP-Exos exert more significant effects on tendon-to-bone healing than normal BMSC-derived exosomes (BMSC-Exos). Here, we isolated and characterized the two types of exosomes, conducted in vitro experiments to measure their effects on fibroblasts (NIH3T3), and performed in vivo experiments to compare the effects on tendon-to-bone integration. Moreover, functional exploration of exosomal miRNAs was further performed by utilizing a series of gain- and loss-of-function experiments. Experimental results showed that both BMSC-Exos and IONP-Exos could be shuttled intercellularly into NIH3T3 fibroblasts and enhanced fibroblast activity, including proliferation, migration, and fibrogenesis. In vivo, we found that IONP-Exos significantly prevented peri-tunnel bone loss, promoted more osseous ingrowth into the tendon graft, increased fibrocartilage formation at the tendon-bone tunnel interface, and induced a higher maximum load to failure than BMSC-Exos. Furthermore, overexpression of miR-21-5p remarkably enhanced fibrogenesis in vitro, and SMAD7 was shown to be involved in the promotive effect of IONP-Exos on tendon-to-bone healing. Our findings may provide new insights into the regulatory roles of IONPs in IONP-Exos communication via stimulating exosomal miR-21-5p secretion and the SMAD7 signaling pathway in the fibrogenic process of tendon-to-bone integration. This work could provide a new strategy to promote tendon-to-bone healing for tissue engineering in the future.

Validation of CT-Based Three-Dimensional Preoperative Planning in Comparison with Acetate Templating for Primary Total Hip Arthroplasty.

OBJECTIVE: This study aims to compare the accuracy of CT-based preoperative planning with that of acetate templating in predicting implant size, neck length, and neck cut length, and to evaluate the reproducibility of the two methods. METHODS: This prospective study was conducted between August 2020 and March 2021. Patients who underwent elective primary total hip arthroplasty by a single surgeon were assessed for eligibility. The included patients underwent both acetate templating and CT-based planning by two observers after the operation. Each observer conducted both acetate templating and CT-based planning twice for each case. The outcome measures included the following: (1) the accuracy of surgical planning in predicting implant size, calcar length, and neck length, which was defined as the difference between the planned size and length and the actual size and length; (2) reproducibility of the two planning techniques, which were assessed by inter-observer and intra-observer reliability analysis; (3) the influence of potential confounding factors on planning accuracy, which was evaluated using generalized estimating equations. RESULTS: A total of 57 cases were included in the study. CT-based planning was more accurate than acetate templating for predicting cup size (93% vs 79%, p < 0.001) and stem size (93% vs 75%, p < 0.001). When assessed by mean absolute difference, the comparison between acetate templating and CT-based planning was 4.28 mm vs 3.74 mm (p = 0.122) in predicting neck length and 3.05 mm vs 2.93 mm (p = 0.731) in predicting neck cut length. In the inter-observer reliability analysis, an intraclass correlation coefficient (ICC) of 0.790 was achieved for predicting cup size, and an ICC of 0.966 was achieved for predicting stem size using CT-based planning. In terms of intra-observer reliability, Observer 1 achieved an ICC of 0.803 for predicting cup size and 0.965 for predicting stem size in CT-based planning. Observer 2 achieved ICC values of 0.727 and 0.959 for predicting cup and stem sizes, respectively. The average planning time was 6.48 ± 1.55 min for CT-based planning and 6.12 ± 1.40 min for acetate templating (p = 0.015). CONCLUSION: The CT-based planning system is more accurate than acetate templating for predicting implant size and has good reproducibility in total hip arthroplasty.

Identification of Novel FBN2 Variants in a Cohort of Congenital Contractural Arachnodactyly.

Congenital contractural arachnodactyly (CCA) is a rare autosomal dominant disorder of connective tissue characterized by crumpled ears, arachnodactyly, camptodactyly, large joint contracture, and kyphoscoliosis. The nature course of CCA has not been well-described. We aim to decipher the genetic and phenotypic spectrum of CCA. The cohort was enrolled in Beijing Jishuitan Hospital and Peking Union Medical College Hospital, Beijing, China, based on Deciphering disorders Involving Scoliosis and COmorbidities (DISCO) study (http://www.discostudy.org/). Exome sequencing was performed on patients' blood DNA. A recent published CCA scoring system was validated in our cohort. Seven novel variants and three previously reported FBN2 variants were identified through exome sequencing. Two variants outside of the neonatal region of FBN2 gene were found. The phenotypes were comparable between patients in our cohort and previous literature, with arachnodactyly, camptodactyly and large joints contractures found in almost all patients. All patients eligible for analysis were successfully classified into likely CCA based on the CCA scoring system. Furthermore, we found a double disease-causing heterozygous variant of FBN2 and ANKRD11 in a patient with blended phenotypes consisting of CCA and KBG syndrome. The identification of seven novel variants broadens the mutational and phenotypic spectrum of CCA and may provide implications for genetic counseling and clinical management.

Development and Validation of an Artificial Intelligence Preoperative Planning System for Total Hip Arthroplasty.

Background: Accurate preoperative planning is essential for successful total hip arthroplasty (THA). However, the requirements of time, manpower, and complex workflow for accurate planning have limited its application. This study aims to develop a comprehensive artificial intelligent preoperative planning system for THA (AIHIP) and validate its accuracy in clinical performance. Methods: Over 1.2 million CT images from 3,000 patients were included to develop an artificial intelligence preoperative planning system (AIHIP). Deep learning algorithms were developed to facilitate automatic image segmentation, image correction, recognition of preoperative deformities and postoperative simulations. A prospective study including 120 patients was conducted to validate the accuracy, clinical outcome and radiographic outcome. Results: The comprehensive workflow was integrated into the AIHIP software. Deep learning algorithms achieved an optimal Dice similarity coefficient (DSC) of 0.973 and loss of 0.012 at an average time of 1.86 ± 0.12 min for each case, compared with 185.40 ± 21.76 min for the manual workflow. In clinical validation, AIHIP was significantly more accurate than X-ray-based planning in predicting the component size with more high offset stems used. Conclusion: The use of AIHIP significantly reduced the time and manpower required to conduct detailed preoperative plans while being more accurate than traditional planning method. It has potential in assisting surgeons, especially beginners facing the fast-growing need for total hip arthroplasty with easy accessibility.

Delineation of dual molecular diagnosis in patients with skeletal deformity.

BACKGROUND: Skeletal deformity is characterized by an abnormal anatomical structure of bone and cartilage. In our previous studies, we have found that a substantial proportion of patients with skeletal deformity could be explained by monogenic disorders. More recently, complex phenotypes caused by more than one genetic defect (i.e., dual molecular diagnosis) have also been reported in skeletal deformities and may complicate the diagnostic odyssey of patients. In this study, we report the molecular and phenotypic characteristics of patients with dual molecular diagnosis and variable skeletal deformities. RESULTS: From 1108 patients who underwent exome sequencing, we identified eight probands with dual molecular diagnosis and variable skeletal deformities. All eight patients had dual diagnosis consisting of two autosomal dominant diseases. A total of 16 variants in 12 genes were identified, 5 of which were of de novo origin. Patients with dual molecular diagnosis presented blended phenotypes of two genetic diseases. Mendelian disorders occurred more than once include Osteogenesis Imperfecta Type I (COL1A1, MIM:166200), Neurofibromatosis, Type I (NF1, MIM:162200) and Marfan Syndrome (FBN1, MIM:154700). CONCLUSIONS: This study demonstrated the complicated skeletal phenotypes associated with dual molecular diagnosis. Exome sequencing represents a powerful tool to detect such complex conditions.

Expanding the mutation and phenotype spectrum of MYH3-associated skeletal disorders.

Pathogenic variants in MYH3 cause distal arthrogryposis type 2A and type 2B3 as well as contractures, pterygia and spondylocarpotarsal fusion syndromes types 1A and 1B. These disorders are ultra-rare and their natural course and phenotypic variability are not well described. In this study, we summarize the clinical features and genetic findings of 17 patients from 10 unrelated families with vertebral malformations caused by dominant or recessive pathogenic variants in MYH3. Twelve novel pathogenic variants in MYH3 (NM_002470.4) were identified: three of them were de novo or inherited in autosomal dominant way and nine were inherited in autosomal recessive way. The patients had vertebral segmentation anomalies accompanied with variable joint contractures, short stature and dysmorphic facial features. There was a significant phenotypic overlap between dominant and recessive MYH3-associated conditions regarding the degree of short stature as well as the number of vertebral fusions. All monoallelic variants caused significantly decreased SMAD3 phosphorylation, which is consistent with the previously proposed pathogenic mechanism of impaired canonical TGF-ß signaling. Most of the biallelic variants were predicted to be protein-truncating, while one missense variant c.4244T>G,p.(Leu1415Arg), which was inherited in an autosomal recessive way, was found to alter the phosphorylation level of p38, suggesting an inhibition of the non-canonical pathway of TGF-ß signaling. In conclusion, the identification of 12 novel pathogenic variants and overlapping phenotypes in 17 affected individuals from 10 unrelated families expands the mutation and phenotype spectrum of MYH3-associated skeletal disorders. We show that disturbances of canonical or non-canonical TGF-ß signaling pathways are involved in pathogenesis of MYH3-associated skeletal fusion (MASF) syndrome.

Survey of the demand for care services for older people and the training needs of their care workers: a cross-sectional study in Northeast China.

BACKGROUND: The older population is increasingly utilizing professional healthcare services, while the requirements for caregivers are becoming more demanding. Therefore, it is important to be mindful not only of the service needs of older people but also to consider the training needs of their care workers. The present study aimed to investigate the care service needs for older people and the training needs of their care workers. METHODS: A cross-sectional questionnaire was used to survey 589 residents of 6 nursing homes and 2 geriatric hospitals, 415 medical staff from 7 geriatric hospitals, 5 nursing homes, and 1 community institution, and 372 nursing assistants from 21 nursing institutions in northeast China. RESULTS: The service with the greatest demand and that with which users were most satisfied was regular visits by healthcare personnel, which was the case for 87.27% of the care recipients. Of the medical staff, 75.42% had training needs related to geriatric healthcare, while the most requested training content was the comprehensive assessment of old people. The most requested method for the delivery of training was by self-study online video courses. Of nursing assistants, only 53.4% had obtained the relevant practicing certificate. While 83.6% participated in relevant training, 86% expressed the need for additional training. The majority of this category of staff wished to receive training in everyday care routines, and the majority wanted to learn by way of practical training. CONCLUSIONS: The care needs of the older population are diverse, and the work performed by healthcare personnel is increasing in scope. The existing training system for such care personnel is not perfect, and the demand for training is high. Existing training methods and content require improvement.

PhenoApt leverages clinical expertise to prioritize candidate genes via machine learning.

In recent years, exome sequencing (ES) has shown great utility in the diagnoses of Mendelian disorders. However, after rigorous filtering, a typical ES analysis still involves the interpretation of hundreds of variants, which greatly hinders the rapid identification of causative genes. Since the interpretations of ES data require comprehensive clinical analyses, taking clinical expertise into consideration can speed the molecular diagnoses of Mendelian disorders. To leverage clinical expertise to prioritize candidate genes, we developed PhenoApt, a phenotype-driven gene prioritization tool that allows users to assign a customized weight to each phenotype, via a machine-learning algorithm. Using the ability to rank causative genes in top-10 lists as an evaluation metric, baseline analysis demonstrated that PhenoApt outperformed previous phenotype-driven gene prioritization tools by a relative increase of 22.7%-140.0% in three independent, real-world, multi-center cohorts (cohort 1, n = 185; cohort 2, n = 784; and cohort 3, n = 208). Additional trials showed that, by adding weights to clinical indications, which should be explained by the causative gene, PhenoApt performance was improved by a relative increase of 37.3% in cohort 2 (n = 471) and 21.4% in cohort 3 (n = 208). Moreover, PhenoApt could assign an intrinsic weight to each phenotype based on the likelihood of its being a Mendelian trait using term frequency-inverse document frequency techniques. When clinical indications were assigned with intrinsic weights, PhenoApt performance was improved by a relative increase of 23.7% in cohort 2 and 15.5% in cohort 3. For the integration of PhenoApt into clinical practice, we developed a user-friendly website and a command-line tool.