Genetic spectrum among 2009 Iranian individuals with neuromuscular disorders using next generation sequencing and multiple ligation dependent probe amplification methods

Abstract

Hereditary neuromuscular disorders (NMDs) are clinically and genetically heterogeneous, with variable severity and onset from birth to adulthood. This study retrospectively analyzes genetic findings in 2009 Iranian individuals with suspected NMDs over 11 years to highlight gene involvement and mutational patterns. Patients underwent gene panel sequencing (GPS), whole exome sequencing (WES), or MLPA for PMP22 in cases with suspected Charcot-Marie-Tooth disease type 1 A (CMT1A). The diagnostic yield of GPS/WES was 46%. Dystrophies were the most prevalent, followed by neuropathies and myopathies. The key implicated genes were CAPN3 and DMD for dystrophies; GDAP1 and MME for neuropathies; GNE and ETFDH for myopathies. The most common phenotype group was dystrophies among both individuals with childhood-onset and adulthood-onset, but the most frequent mutated gene was CAPN3 in children and DYSF in adults. Identified variants include 761 (97%) single nucleotide variants (SNVs) and 24 (3%) copy number variations (CNVs). Notably, 26% of SNVs were novel. Among individuals tested with MLPA, 28% had confirmed PMP22 gene deletions or duplications, with 73% being duplications linked to CMT1A. This large-scale analysis provides insight into the genetic landscape of NMDs in Iran. Understanding gene distribution and mutation types can improve diagnosis and inform management strategies for affected individuals.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-26836-x.

Introduction

Inherited neuromuscular disorders (NMDs) are a heterogeneous group of diseases that result from variants in genes necessary for the appropriate functioning of skeletal muscle cells, the peripheral nervous system, motor neurons, or the neuromuscular junction. These disorders vary widely in terms of age of onset, severity, and clinical manifestations, including symptoms such as muscle weakness, dystonia, and neuropathies. Despite their diverse characteristics, a common feature among NMDs is the impairment of voluntary movement, which imposes a significant burden on patients, their families, and society^1,2. Identifying the underlying molecular causes in individuals with NMDs can provide valuable information for diagnosis, prognosis, genetic counseling, and possible targeted therapies. More than 600 genes are associated with 16 different NMD phenotype groups, and more are yet to be discovered³. Moreover, the highly heterogeneous nature of these disorders leaves many patients and families without a genetic diagnosis. Recent advances in next-generation sequencing (NGS), including gene panel sequencing (GPS) and whole exome sequencing (WES), have helped to uncover the genetic cause of many NMD patients.

However, there are some limitations in NGS when it comes to identifying certain types of variations, such as large deletions or duplications. Additionally, for some kinds of NMDs, which are more common and easily recognized, caused by frequent and well-known variants, traditional testing methods like multiple ligation-dependent probe amplification (MLPA) can be beneficial. An example is Charcot-Marie-Tooth disease type 1 A (CMT1A), where deletions or duplications of the PMP22 gene are the main cause⁴, so the MLPA testing is the first-tier approach in these cases.

Increased rates of consanguinity are associated with rare autosomal recessive (AR) inherited diseases due to an increased chance of inheriting homozygous pathogenic variants from both parents. Due to a high consanguinity rate, Iran is a hotspot for discovering novel genes and variants in individuals with rare diseases such as NMDs⁵.

While previous studies have significantly contributed to our understanding of the genetic basis of NMDs, many have focused on specific phenotype groups or limited their cohorts to pediatric populations. This narrow focus leaves a gap in understanding the broader genetic landscape of NMDs across diverse clinical presentations and age groups. To address this, we studied the genetic results of 2009 individuals with suspected NMDs who were referred to our center between 2013 and 2024 by using MLPA, GPS, and WES to gain insight into the genes and mutational variety among the Iranian patients with NMDs.

Results

Demographic profile

A total of 2009 individuals with NMDs were included in this study. There were 1159 males (57.7%), 847 females (42.2%), and 3 fetuses (0.1%) with a median age of 27 years (IQR: 14–39, total range: 0–77).

Diagnostic results of NGS and MLPA methods

A total of 1,538 out of 2,009 individuals suspected of having NMDs, including 76 individuals with negative PMP22 test results, underwent either WES or GPS. The median age at report was 25.5 years (IQR: 14.0–38.0, total range: 0–72). A genetic diagnosis was achieved in 715 individuals (46%). The remaining individuals were classified as follows: 337 with variants of uncertain significance (VUS) (22%), 454 with negative results (30%), and 32 unclear results (2%). Genetic diagnosis rate in different categories is detailed in Table 1.

Table 1.

Summary of studied individuals and diagnostic yield.

Category		Number of cases (%)	Number of diagnosed cases	Diagnostic rate (%)	P-Value
MLPA
Gender	Fetus	3 (0.5)	3	3/3 (100)
	Male	330 (60)	91	91/330 (28)	0.99
	Female	214 (39)	59	59/214 (28)
Age at report	Childhood (< 18 years)	139 (25)	28	28/ 139 (20)	NA*
	Adulthood (≥ 18 years)	405 (74)	122	122/405 (30)
Total		547	153	153/547 (28)
NGS
Gender	Male	872 (57)	395	395/872 (45)	0.28
	Female	666 (43)	320	320/666 (48)
Age at report	Childhood (< 18 years)	513 (33)	273	273/513 (53)	NA*
	Adulthood (≥ 18 years)	1025 (67)	442	442/1025 (43)
Consanguinity	Yes	989 (64)	515	515/989 (52)	< 0.05
	No	549 (36)	200	200/549 (36)
Family history	Yes	414 (27)	232	232/414 (56)	< 0.05
	No	1124 (73)	483	483/1124 (43)
Test	WES	1382 (90)	626	626/1382 (45)	< 0.05
	GPS	156 (10)	89	89/156 (57)
Total		1538	715	715/1538 (46)
Total MLPA and NGS		2009	868	868/2009 (43)

^* NA: Not applicable. The age variable represents age at reporting rather than age at onset; therefore, the p-value is not meaningful for this category.

From 2009, a total of 547 patients in this cohort were primarily diagnosed with CMT and underwent MLPA testing for the PMP22 gene. The median age at report was 31.0 years (IQR: 17.0–41.5, total range: 0–77). Of these, 153 individuals (28%) were confirmed to have either a deletion or a duplication in the PMP22 gene. Among the 153 cases, duplications were identified in 112 (73%) with CMT1A (OMIM: 118220), and deletions were identified in 41 (27%) with Hereditary neuropathy with liability to pressure palsy (HNPP) (OMIM: 162500) (see Table 1).

The subsequent results presented below refer primarily to the individuals who underwent NGS analysis (WES or GPS), excluding those diagnosed solely via MLPA.

The most frequent mutated genes and phenotype grouping of diagnosed individuals

Positive results were observed in 167 different NMD genes, with the most prevalent being CAPN3 (n = 60), DYSF (n = 43), DMD (n = 37), SPG11 (n = 19), LAMA2 (n = 18), CHRNE (n = 17), SGCA (n = 16), CLCN1 (n = 15), GNE (n = 14), SOD1 (n = 14), RYR1 (n = 12), ETFDH (n = 11), ATM (n = 11), GDAP1 (n = 10), LMNA (n = 10), MME (n = 10), COL6A2 (n = 9), SACS (n = 9), SORD (n = 8), COLQ (n = 8) accounting for almost half of all diagnosed genes. (Supplementary Table 2)

Among the 81 genes recommended by the American College of Medical Genetics and Genomics (ACMG) for reporting of incidental findings due to their potential clinical actionability, we identified pathogenic or likely pathogenic variants in 11 such genes in 46 individuals from our cohort. The affected genes included LMNA, RYR1, MYH7, CACNA1S, TTN, BAG3, FLNC, DES, GAA, TPM1, and TTR.

After analyzing 715 individuals with confirmed positive results, we categorized them into nine different groups, including dystrophies, neuropathies, myopathies (other than dystrophies), neuromuscular junction disorders (NMJs), motor neuron diseases (MNDs), hereditary paraplegia, hereditary ataxia, and skeletal muscle ion channelopathies (SMICs) based on their phenotypes and their corresponding mutated gene. Notably, 86 individuals were categorized into overlapping groups, as their phenotype spanned two or more phenotypic groups. Dystrophies constituted the largest category (34%), followed by neuropathies (14%) and myopathies (14%) (Fig. 1). Among individuals with VUS results, overlapping phenotypes (29%) were the most prevalent, followed by dystrophies (19%) and neuropathies (15%).

Fig. 1.

The percentage of individuals in each NMD phenotypic category.

Moreover, recurrently mutated genes in each group are studied (Fig. 2). The most frequently mutated genes in each category are as follows; CAPN3 in dystrophies (27%), GDAP1 in neuropathies (9%), GNE in myopathies (14%), CHRNE in NMJs (37%), SOD1 in MNDs (35%), SPG11 in hereditary paraplegia (30%), ATM in hereditary ataxias (28%), and CLCN1 in SMICs (100%).

Fig. 2.

The most recurrently mutated genes across different NMD phenotypic categories.

Among 76 individuals whose MLPA testing for the PMP22 gene was negative and who were referred for NGS, 34 were found to have positive results. Most cases were categorized into the neuropathies group (32, 94%) with MPZ (5, 15%), GJB1 (4, 12%), SH3TC2 (4, 12%), and PRX (4, 12%) genes being the most recurrent ones in these groups. One individual was categorized into MNDs as the mutated gene was ASAH1, and one individual into hereditary ataxias with SACS being the mutated gene.

Phenotypic grouping and gene distribution between childhood and adulthood-onset individuals

The age of onset was reported for 665 of the 715 individuals with positive results, with 476 (72%) exhibiting childhood onset (under 18 years of age) and 189 (28%) showing adulthood onset (age 18 years and above). Among individuals with childhood-onset NMDs, dystrophies were the most common group (42%), while MNDs were the least common (3%). Similarly, among individuals with adulthood-onset NMDs, dystrophies were the most common phenotype (27%), while NMJs were the least common (3%). Notably, SMICs were not reported among the adulthood-onset group (Fig. 3). The most frequently mutated gene in children was CAPN3 (9%), whereas in adults, it was DYSF (13%) (Fig. 4).

Fig. 3.

The phenotypic distribution among individuals with childhood onset and individuals with adulthood onset.

Fig. 4.

The distribution of mutated genes among children and adults. Figure 4a The distribution of mutated genes in individuals with adulthood-onset phenotypes. Figure 4b The distribution of mutated genes in individuals with childhood-onset phenotypes.

Mutation spectrum among individuals with NMDs using NGS

In total, 784 variants, including SNVs and CNVs, were identified across 167 NMD genes, with 432 classified as likely pathogenic (55%), 330 as pathogenic (42%), and 22 as VUS (3%). Notably, individuals with VUS variants in a compound heterozygous state, along with a pathogenic or likely pathogenic variant, were considered positive in their final results. Among 761 SNVs, missense variants were the most common (41%), followed by frameshift variants (22%), stop gain (20%), splicing (11%), and Other types of variants, including 5’ UTR, inframe deletion, inframe duplication, initiation, stop loss, and synonymous (3%). Of all SNVs, 199 (26%) were novel and had not been previously reported. Figure 5 presents the number of different mutation types among the most frequently mutated genes in this study. CNVs were seen in 13 different genes with 21 deletions (88%) and SGCB (35%), SGCG (17%), and DYSF (9%) genes being the most prevalent ones. Other Genes with CNVs include AGL, PMP22, DCTN1, SPG7, SPG21, LAMA2, DMD, LPIN1, and DDHD1. Most variants were in a homozygous state (61%), followed by compound heterozygous (18%), with an autosomal recessive mode of inheritance. Other types of inheritance are autosomal dominant (14%) and X-linked (7%). Among the 108 variants inherited in an autosomal dominant manner, the segregation status of 47 individuals was examined, revealing that 29 were de novo variants.

Fig. 5.

The distribution of different variation types among the most prevalent mutated NMD genes. “Other” includes 5′ UTR, in-frame deletion, in-frame duplication, initiation, stop-loss, and synonymous variants.

Discussion

Diagnostic yield

NMDs, as clinically and genetically heterogeneous conditions, pose significant diagnostic challenges. Choosing appropriate molecular testing is necessary for overcoming these challenges. In recent years, individuals with a wide range of NMD phenotypes and ages have been particularly understudied in regions like Iran, which have high rates of consanguineous marriage. To fill this gap, we retrospectively studied the results of 547 CMT1A individuals referred for MLPA and 1538 individuals referred for GPS and WES tests. Our study shows the comprehensive genetic and phenotypic diversity of hereditary NMDs with the help of both traditional and advanced molecular diagnostic methods of MLPA and NGS in a large cohort of Iranians with high rates of consanguinity. The 46% diagnostic yield of NMDs using NGS in our study is consistent with previous studies of NMD diagnosed with NGS^6–11.

Combined testing

This study, with a 28% diagnostic yield of MLPA, emphasizes the fact that some traditional tests could still be used as a complementary method to NGS for well-known diseases like CMT1A. Particularly, MLPA is an effective method to identify large deletions and duplications that are challenging to identify by NGS. We should not underestimate the effect of NGS as a complementary test for MLPA. For example, MLPA results for PMP22 deletion or duplication were negative for two individuals initially diagnosed with CMT (Case232 and Case516). However, NGS later identified pathogenic SNVs in the same gene. Moreover, in the recently published paper on spinal muscular atrophy (SMA), 3.63% (47/1295) of individuals with a SMA diagnosis had only one copy deletion of SMN1, which underscores the need for NGS to identify the mutation in the remaining copy of this gene in these individuals¹². For instance, in our cohort, one individual diagnosed with SMA, Case195, MLPA testing for SMN1 deletions revealed only one copy of exon7/8 of the SMN1 gene. Subsequent WES identified a Loss-of-function (LOF) variant in the remaining copy. This was the only case with a mutation in the SMN1 gene in our cohort, and both NGS and MLPA testing provided a final diagnosis for this case. A similar case was also seen in another study, an individual presenting with SMA, exhibited compound heterozygous variations (c.549del; p.Lys184Serfs∗and exon 7 and 8 deletion) in the SMN1 gene¹³. These observations highlight that a combined approach of diagnostic tests is necessary in neuromuscular disorders, as they are genetically heterogeneous conditions.

Consanguinity

In regions characterized by a notably high rate of consanguineous marriage, the possibility of rare conditions due to increased risk of inheritance of pathogenic variants in a homozygous state is elevated^5,14. Our results align with the same observations; comparing the consanguineous and non-consanguineous families reveals the diagnostic rate of 52% and 36%, respectively (Table 1). Furthermore, 79% of individuals with positive results were affected by an autosomal recessive mode of inheritance, either by homozygous or compound heterozygous variants. These findings highlight the need to adapt genetic screening and counseling to populations with a high rate of consanguinity. In such settings, carrier screening before or during marriage could help identify couples at risk of having children with genetic conditions. Genetic counsellors should also consider the higher chance of recessive disorders and focus testing on detecting homozygous and compound heterozygous variants. These tailored approaches can improve diagnosis and support informed family planning decisions.

Mutation spectrum and CNVs

A large cohort provides an opportunity to see the pattern of the pathogenicity that differs from one gene to another. For example, the majority of pathogenic variants in DYSF, DMD, LAMA2, and SPG11 are LOF variants that can affect and disrupt the expression or translation of a gene into a functional protein, while the majority of pathogenic variants in SGCA, GNE, SOD1, ETFDH, and LMNA are missense variants that can affect the function of the protein in the cell. Additionally, the larger size of the genes increases the chance of pathogenic variants, and when coupled with an escalated AR inheritance pattern, this contributes to the elevated incidence of novel variants. In our cohort, this trend is obvious, especially in four large genes like DYSF (6%), ATM (6%), DMD (4%), and NEB (4%), with the largest proportion of identified novel variants. (see Supplementary Table 2)

In this cohort, 23 CNVs in 13 different genes were also detected; 8 of which were only detected in the SGCB gene, and all of them were deletions of exon 2. In 2017, Alavi et al. and in 2020, Mojbafan et al. reported that the deletion of exon 2 of this gene could be a founder effect, which is seen mostly in southern regions of Iran^15,16. In Case91, compound heterozygous variants in the DYSF gene were identified, comprising SNV (c.4639-1G >A) and CNV (exons 12–13 deletion). This can emphasize the importance of finding other strong heterozygous variants, either CNV or SNV, in the non-coding region of the same gene where only a single pathogenic or likely pathogenic variant is initially identified in AR genes. These variants can be identified by analyzing CNVs through WES or performing whole-genome sequencing (WGS) in the non-coding region of the gene. Other studies also reported similar cases where a heterozygous SNV variant was detected by NGS testing, and subsequently, further methods for CNV detection could lead to the identification of a deletion or duplication in the same gene^17–19. In our cohort, we identified 17 pathogenic or likely pathogenic heterozygous SNVs in genes with AR inheritance that require more investigation, and all were checked for any CNV in the same gene; only case91 has a CNV in the same gene. Accordingly, these variants were analyzed in both parents using Sanger sequencing and MLPA, revealing that each variant was inherited from a different parent. This highlights the value of combining SNV and CNV analysis, particularly in AR disorders where a single heterozygous variant may initially be detected. Therefore, when only a single heterozygous variant is identified in an AR gene, it may be beneficial to consider CNV analysis as part of the diagnostic workflow. This integrated approach has the potential to improve diagnostic yield and help identify clinically relevant compound heterozygous cases that might otherwise be missed.

Phenotypic groups

Categorizing individuals with definitive diagnoses into nine different groups revealed that the most frequent phenotypic groups were dystrophies, neuropathies, and myopathies, and the least frequent were hereditary ataxia, hereditary paraplegia, and SMICs, which almost align with the prevalence of NMDs when analyzing the results of individuals who underwent NGS test^6,7. Nevertheless, MNDs are among the most frequent NMD conditions worldwide. For example, amyotrophic lateral sclerosis (ALS) and SMA are estimated to affect 4.1–8.4/100,000 and 1–2/100,000 individuals, respectively^20,21. In our cohort, however, MNDs rank sixth out of nine groups. This is primarily because common variants in this group, such as repeat expansions in C9orf72 for ALS and exon 7/8 deletions in SMN1 for SMA, are assessed using alternative testing methods in our referral center and were not included in the current sequencing cohort. In addition, SMA cases, which represent a substantial portion of MNDs in our cohort, were recently analyzed and reported in a separate publication¹².

In dystrophies, the most frequent genes were related to Limb Girdle muscular dystrophy (LGMD) and Duchenne muscular dystrophy (DMD). Although, DMD is the most common NMD globally with prevalence of 2.8–4.8/100,000^22,23, in our cohort, it stands in third place of frequent genes in this group which is due to the fact that those patients diagnosed with DMD are referred for MLPA testing of DMD gene deletion as a first-tier test and data related to these groups of NMD patients are included in another paper which is in progress. The most frequent genes in LGMD-related individuals were CAPN3, DYSF, LAMA2, SGCA, and SGCB. This may differ slightly in other studies based on the objective population. For instance, in Austria, the most prevalent mutated genes of LGMD were reported as CAPN3, FKRP, ANO5, DYSF, and SGCA²⁴. In the United States, the most frequently mutated genes in LGMD are reported as CAPN3, DYSF, GAA, ANO5, and FKRP²⁵.

In neuropathies, the most frequently mutated genes after PMP22 duplication are GJB1, MFN2, and MPZ²⁶. However, in our cohort, this trend was slightly different in the neuropathies group, with GDAP1, MME, MFN2, RRX, and SH3TC2 being the most frequently mutated genes. In individuals who are negative for PMP22 deletion or duplication, the most prevalent genes identified were MPZ, GJB1, SH3TC2, and PRX, which closely align with variants reported in similar patient cohorts²⁷. CMTX6 is a considerably rare condition caused by PDK3, and so far, only four families from Australia, Korea, Germany, and Brazil have been reported with variants in this gene^28–31. In our cohort, we identified a homozygous variant of c.473G >A (p.Arg158His) in the PDK3 gene. Interestingly, this variant is the same as reported in Australian and Korean families.

The large number of the “overlapping group” shows a significant limitation of this study. Particularly, the overlapping phenotypes have been noted in genes like DYSF, LMNA, and ANO5, which are related to more than one phenotype as represented in OMIM. This highlights the necessity of additional clinical data, such as EMG, CT scan, MRI, and muscle biopsy, to facilitate the precise diagnosis of a single phenotype. These findings emphasize the importance of the complementary role of clinicians and geneticists in achieving a definitive and accurate diagnosis.

Phenotypic distribution by age

The severity of phenotypes associated with NMDs and the age of manifestations can differ significantly based on the type of involved variants and genes. Subsequently, the distribution of genes can vary significantly among children and adults. To our current knowledge, we are the first to compare the distinct pattern of gene distribution among children and adults diagnosed with NMDs. The most frequently mutated genes in children were CAPN3, DMD, LAMA2, DYSF, CHRNE, CLCN1, and SPG11, while in adults were DYSF, CAPN3, SOD1, GNE, ETFDH, MME, and HSPB1. This implies that some genes, like DYSF and CAPN3, can cause early-onset or late-onset disease based on the type of mutation, while some genes can cause a severe form of NMDs. This insight provides multiple advantages for disease management. Firstly, clinicians can offer a better diagnostic approach by prioritizing specific genes. Secondly, genetic counseling can guide families more precisely based on the age of onset of phenotypes.

Among our cohort, 46 individuals carried pathogenic or likely pathogenic variants in 11 genes listed by the ACMG as medically actionable. These genes are associated with conditions where early detection, surveillance, and intervention can significantly improve clinical outcomes³². Identifying such variants not only benefits the individuals tested but also has important implications for their family members, who may carry the same variant and be eligible for preventive care through cascade testing. This highlights the broader utility of genomic screening beyond primary diagnoses and emphasizes the importance of structured follow-up, including genetic counseling and family-based risk assessment.

Implications and limitations

To the best of our knowledge, this is the largest cohort of individuals with NMDs from regions with a high rate of consanguinity, demonstrating the genetic diversity of these conditions. Despite achieving diagnosis in 46% of NMD individuals referred to NGS in our cohort, there are still inconclusive cases that had been found to have uncertain variant of significance (VUS cases, 22%), in whom functional and/or segregation analysis should be used to confirm or reject the pathogenicity, or no causative variant (Negative cases, 30%) detected. This reflects the challenges and limitations of the diagnosis process. One factor is related to our role as a referral diagnostic center. Some clinical data, such as MRI, CT scan, muscle biopsy, and EMG, which are crucial for differential diagnosis and accurately interpreting genetic results, are not provided for some individuals. Another factor is due to the limitations of NGS in identifying specific variants, for example, CNVs and repeat expansion variants. These variants play an important role in the pathogenicity of some NMDs and require other methods like QF-PCR or MLPA to be identified. Moreover, 2% unclear results that individuals were carriers of pathogenic or likely pathogenic variants in genes with an AR pattern of inheritance emphasize the need for other methods in order to identify the next variant in the same gene.

Furthermore, it is essential to recognize that some individuals with negative genetic results may still have a potentially treatable neuromuscular disorder. For instance, anti-HMGCR myopathy, an immune-mediated necrotizing myopathy that can clinically mimic limb-girdle muscular dystrophy (LGMD) phenotypes, has been reported³³. This condition exemplifies how accurate diagnosis fundamentally alters patient management and treatment strategies, as affected individuals can respond to immunosuppressive therapy. Such cases highlight the need to integrate molecular findings with detailed clinical phenotyping and histopathological evaluation to avoid misclassifying treatable myopathies as genetic LGMD forms.

To address the limitations observed in negative and VUS cases, several strategies can be employed. Reanalysis of NGS data over time, as variant databases and interpretation guidelines evolve, may lead to variant reclassification. Whole genome sequencing (WGS) can help detect structural variants, non-coding variants, or deep intronic variants missed by exome sequencing. Transcriptome (RNA) sequencing may provide insights into splicing defects or gene expression changes, especially in genes with uncertain variants. Functional studies, including in vitro or in vivo assays and segregation analysis in families, are essential to establish the pathogenicity of VUS. Together, these advanced approaches can increase diagnostic yield, refine variant interpretation, and ultimately enhance patient care in genetically complex disorders such as NMDs³⁴.

In conclusion, our large cohort of 2009 individuals with NMDs expands the knowledge of the genetic diversity of neuromuscular disorders in the Iranian population. This overview could offer benefits such as information about the risks associated with consanguineous marriage, including an increased risk of having a child affected by a rare NMD disease. It also provides a valuable resource for genetic counseling and preventive healthcare planning, can guide diagnostic testing strategies, and optimize clinical management.

Methods and materials

Study population

In the past 11 years (2013–2024), 2,009 individuals were referred by their clinicians to the Molecular Genetic Department at the Kariminejad – Najmabadi Pathology & Genetic Center in Tehran, Iran, for genetic analysis of NMDs. We started using the NGS method in 2013, and since then, we have included all males and females of any age with suspected NMDs who were referred to NGS testing and all CMT individuals who were referred to MLPA testing in this study. Individuals without a neuromuscular diagnosis were excluded. The external clinicians provided all clinical assessments and phenotype information. Our center functioned as a referral diagnostic laboratory, and our in-house clinicians conducted genotype-phenotype correlations based on provided clinical data and sequencing results. In the case of discrepancy or unclear result, reanalysis of genetic data or additional testing strategies were recommended. Further detailed information on individuals referred to NGS is provided in Supplementary Table 1.

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the University of Social Welfare and Rehabilitation Sciences, Tehran, Iran (IR.USWR.REC.1404.081). Written informed consent was obtained from each patient or their guardian. The study was conducted in accordance with the Declaration of Helsinki.

Sample collection

Peripheral blood samples from index cases and available family members, as well as fetal tissue from autopsies, were collected for genomic DNA extraction. Genetic counseling was provided both before and after the test by genetic counselors and/or clinical geneticists. Each individual’s clinical data, genetic test history, and family pedigree were carefully reviewed and documented.

Multiplex ligation-dependent Probe Amplification (MLPA)

MLPA for PMP22 gene testing was conducted on 547 individuals with suspected Charcot-Marie-Tooth disease type 1 A (CMT1A) using SALSA MLPA Probemix P033-B4 CMT1 test kit (MRC Holland, Amsterdam, The Netherlands; http://www.mrc-holland.com) and following the manufacturer’s protocols. Amplification products were identified and quantified by capillary electrophoresis on an ABI 3130 genetic analyzer (Applied Biosystems, Foster City, CA, USA) and were analyzed using the Cofalyser.Net™ V220513.1739 Software (MRC Holland, Amsterdam, The Netherlands).

Gene panel sequencing (GPS), Whole-exome sequencing (WES), and data analysis

GPS or WES was performed on 1,538 individuals with suspected NMDs (including 76 negative results for PMP22 gene testing) (Fig. 6). In the case of GPS, the genes that are included in each gene panel are provided in Supplementary Table 3. Notably, the use of GPS or WES for individuals was based on requests from external clinicians. The majority of cases underwent singleton-based sequencing (99%), with trio-based sequencing performed in only 1% of cases. About 99.8% of the target regions were covered at ≥ 10× read depth and 99.1% at ≥ 40×, with a mean coverage depth of 100X. Library preparation, sequencing, and analysis of variants were performed as described in a previously published paper (This study also included data from 548 cases that were previously reported in that publication, and these 548 cases are included in the dataset)⁵.

Fig. 6.

Number of individuals referred to Kariminejad – Najmabadi Pathology and Genetics Center in an 11-year period for different methods of neuromuscular testing.

In addition to classifying results as positive (individuals with pathogenic or likely pathogenic variants), VUS (individuals with uncertain significance variants), or negative (individuals with no identified variant), we introduced a fourth category: “unclear results.” This category represented cases with findings of potential clinical relevance that did not meet the criteria for a definitive genetic diagnosis. Specifically, “unclear results” included:

• Heterozygous variants in AR genes where the patient’s phenotype closely matched the associated OMIM phenotype, but a second pathogenic or likely pathogenic variant was not identified. In such cases, a second variant may have been missed due to technical limitations of exome sequencing (e.g., undetected CNVs, deep intronic variants, or low coverage regions).

• Variants that partially explained the clinical phenotype, while other unexplained features suggested the possibility of additional variants in other genes contributing to the overall presentation.

Variant calling for copy number analyses utilized the Illumina Dragen genome-wide depth-based CNV caller. Annotation of CNVs was performed using AnnotSV, an integrated online tool for structural variations annotation³⁵. For filtering common CNVs, gnomAD, DGV (Database of Genomic Variants) (http://dgv.tcag.ca), and dbVar (https://www.ncbi.nlm.nih.gov/dbvar/) were employed. The final candidate variants were classified following the American College of Medical Genetics and Genomics/Association for Molecular Pathology (ACMG/AMP) guidelines and following the Clinical Genome (ClinGen) recommendations for utilizing ACMG/AMP criteria³⁶. With guidance from the relevant recommendations and classification guidelines, final variant classifications were made by an in-house expert committee.

Sanger sequencing and segregation analysis

Candidate variants detected by NGS with low coverage were confirmed by PCR amplification and Sanger sequencing. Segregation analysis was performed for variants of VUS and candidate pathogenic or likely pathogenic variants when requested by the referring clinician and when samples from family members were available.

Following diagnosis, individuals were categorized into groups based on an online neuromuscular gene table and their phenotypes³. Diagnosed individuals were also classified into two different age of onset groups, including childhood (under 18 years old) and adulthood (18 years and older), based on the age of onset of the phenotype’s manifestations.

Statistical analyses

All statistical analyses were performed using R Studio software, version 2023.09.1 + 494, and Microsoft Excel. Descriptive statistics and data visualization were primarily conducted in Excel, which was also used to generate all tables presented in the study. Comparisons of diagnostic rates between groups were assessed using the Pearson chi-square test. A p-value of < 0.05 was considered statistically significant.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

H.N. initiated and led the project, providing financial support with contributions from N.M., who participated in the study’s design, performed data analysis, reviewed the literature, and drafted the manuscript. Contributions were also made by P.A. and A.K., who were involved in data collection, literature review, data analysis, and manuscript drafting. M.B. coordinated data collection, provided clinical interpretation of molecular findings, and supported statistical analysis. F.A. developed and supervised the next-generation sequencing bioinformatics pipeline, performed sequencing data analysis, and contributed intellectual input to the manuscript. AR.K. clinically evaluated the results and supervised the genetic counselling team. K.N., F.F., S.Z.N., F.AF., and H.Y. contributed to the genetic counselling of patients. S.D., M.F., M.A., Z.G., Z.E., R.V., E.P., M.M.N., P.S., and A.A. performed sequencing data analysis, participated in data collection and reevaluation, and were involved in segregation and sequencing data analysis. M.A.K. and M.AZ. conducted MLPA testing and analysis. F.P. was responsible for patient recruitment and evaluation. S.M.K., M.R., F.Z., H.S., S.A., MK.B., A.F., G.Z., M.R.A., P.J., P.S., A.A.O., B.H.A., F.FA., P.K., S.N., K.K., and A.K. were responsible for patient recruitment and contributed clinical expertise regarding the patients’ details. All authors read and approved the final manuscript.

Funding

This research received no external funding.

Data availability

All relevant data are included in this published article and its supplementary information files. Additional datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. All identified variants in this study have been submitted to the ClinVar database under Submission IDs of SUB15283860 and SUB15323776, and accession numbers are provided in Supplementary Table 3.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the University of Social Welfare and Rehabilitation Sciences, Tehran, Iran (IR.USWR.REC.1404.081). Written informed consent was obtained from each patient or their guardian. The study was conducted in accordance with the Declaration of Helsinki.

Competing interests

The authors declare no competing interests.