An evaluation of statistical procedures for comparing an individual's performance with that of a group of controls

The single case methodology that is widely used in cognitive neuropsychology often requires a comparison of data from a single individual (the patient) with that from a group of controls, in order to ascertain whether the patient's mean score can be viewed as significantly different from that of controls. This article reviews methods that have been used to deal with such data. Although Analysis of Variance (ANOVA) provides one possible solution of comparing group means, unequal group sizes and differences in variability between patient and controls can violate the assumptions of the test. Using Monte Carlo simulations, it was found that differences in group size and a group of N = 1 did not significantly affect the reliability of the analysis. In contrast, unacceptably high Type I errors were obtained when, in addition to unequal group sizes, there were relatively modest differences between the variance of the patient and that of the controls. We suggest that ANOVA can be used for the comparison of the mean score of an individual with that of a group of controls, but that when there is a difference in variability between the two groups, revised F criteria should be used in order to make the analysis reliable. A table of modified F values is given, which can be used for various departures from homogeneity of variance.     

An evaluation of statistical procedures for comparing an individual's performance with that of a group of controls

The single case methodology that is widely used in cognitive neuropsychology often requires a comparison of data from a single individual (the patient) with that from a group of controls, in order to ascertain whether the patient's mean score can be viewed as significantly different from that of controls. This article reviews methods that have been used to deal with such data. Although Analysis of Variance (ANOVA) provides one possible solution of comparing group means, unequal group sizes and differences in variability between patient and controls can violate the assumptions of the test. Using Monte Carlo simulations, it was found that differences in group size and a group of N = 1 did not significantly affect the reliability of the analysis. In contrast, unacceptably high Type I errors were obtained when, in addition to unequal group sizes, there were relatively modest differences between the variance of the patient and that of the controls. We suggest that ANOVA can be used for the comparison of the mean score of an individual with that of a group of controls, but that when there is a difference in variability between the two groups, revised F criteria should be used in order to make the analysis reliable. A table of modified F values is given, which can be used for various departures from homogeneity of variance.     

Genetic evidence and integration of various data sources for classifying uncertain variants into a single model

Genetic testing often results in the finding of a variant whose clinical significance is unknown. A number of different approaches have been employed in the attempt to classify such variants. For some variants, case-control, segregation, family history, or other statistical studies can provide strong evidence of direct association with cancer risk. For most variants, other evidence is available that relates to properties of the protein or gene sequence. In this work we propose a Bayesian method for assessing the likelihood that a variant is pathogenic. We discuss the assessment of prior probability, and how to combine the various sources of data into a statistically valid integrated assessment with a posterior probability of pathogenicity. In particular, we propose the use of a two-component mixture model to integrate these various sources of data and to estimate the parameters related to sensitivity and specificity of specific kinds of evidence. Further, we discuss some of the issues involved in this process and the assumptions that underpin many of the methods used in the evaluation process.     

Non-invasive screening method for Fabry disease by measuring globotriaosylceramide in whole urine samples using tandem mass spectrometry

Fabry disease is an X-linked sphingolipidosis due to a deficiency of alpha-galactosidase A, which leads to the accumulation of globotriaosylceramide (GL-3) in several organs. When recombinant human alpha-galactosidase A is intravenously administered repeatedly before the patient develops permanent tissue damage, there is evidence that the accumulation of GL-3 is decreased in some organs and that the clinical symptoms are alleviated in some patients. However, Fabry disease is rare and many patients are not diagnosed until adulthood after irreversible tissue damage has occurred. Our group has developed a simple and non-invasive screening method for Fabry disease that measures total GL-3 in whole urine samples by tandem mass spectrometry. Using this method, we found that the concentration of GL-3 in whole urine sample from hemizygous patients, including pre-symptomatic young children with classic type Fabry disease, was significantly higher than that in controls. The mean concentration of GL-3 in urine from heterozygotes with symptoms was significantly higher than control concentrations, but GL-3 levels in the urine from 2 out of 8 heterozygotes of classic type Fabry disease were within control levels. An asymptomatic 14-year old hemizygote in the family of a cardiac variant did not have elevated urinary GL-3. Therefore, screening for the classic type and probably renal variant of Fabry disease is possible by measuring urinary GL-3, using our method. The early diagnosis of cardiac variant hemizygotes and some heterozygotes with all types of Fabry disease will not be possible using our method. We propose that this procedure can be used as a reliable, non-invasive, simple method for general and high-risk population screening for hemizygotic patients with the classic type and probably renal variant of Fabry disease.     

一例多种羧化酶缺乏症的HLCS基因变异分析

目的对1例临床诊断为多种羧化酶缺乏症(multiple carboxylase deficiency,MCD)的患儿及其父母进行相关致病基因的变异分析,为临床诊断及遗传咨询提供依据。方法应用PCR技术和DNA测序技术对患儿的MCD致病基因BT和HLCS编码区进行变异检测,并对患儿父母进行相应基因变异分析。在80名正常人中对未报道过的基因变异进行PCR-限制性片段长度多态性分析。结果患儿的BT基因编码区未发现碱基改变,HLCS基因存在c.286delG(p.Val96Leufs*162)和c.1648G>A(p.Val550Met)复合杂合变异,其中c.286delG(p.Val96Leufs*162)经PCR-限制性片段长度多态性分析验证为新变异。结论HLCS基因c.286delG(p.Val96Leufs*162)和c.1648G>A(p.Val550Met)变异可能为患儿的致病原因,致病基因的检出为临床诊断及遗传咨询提供了依据,同时丰富了HLCS基因变异谱。     

In silico analysis of missense substitutions using sequence-alignment based methods

Genetic testing for mutations in high-risk cancer susceptibility genes often reveals missense substitutions that are not easily classified as pathogenic or neutral. Among the methods that can help in their classification are computational analyses. Predictions of pathogenic vs. neutral, or the probability that a variant is pathogenic, can be made based on: 1) inferences from evolutionary conservation using protein multiple sequence alignments (PMSAs) of the gene of interest for almost any missense sequence variant; and 2) for many variants, structural features of wild-type and variant proteins. These in silico methods have improved considerably in recent years. In this work, we review and/or make suggestions with respect to: 1) the rationale for using in silico methods to help predict the consequences of missense variants; 2) important aspects of creating PMSAs that are informative for classification; 3) specific features of algorithms that have been used for classification of clinically-observed variants; 4) validation studies demonstrating that computational analyses can have predictive values (PVs) of approximately 75 to 95%; 5) current limitations of data sets and algorithms that need to be addressed to improve the computational classifiers; and 6) how in silico algorithms can be a part of the "integrated analysis" of multiple lines of evidence to help classify variants. We conclude that carefully validated computational algorithms, in the context of other evidence, can be an important tool for classification of missense variants.     

Nonsense Mutation in Coiled‐Coil Domain Containing 151 Gene (CCDC151) Causes Primary Ciliary Dyskinesia

Primary ciliary dyskinesia (PCD) is an autosomal‐recessive disorder characterized by impaired ciliary function that leads to subsequent clinical phenotypes such as chronic sinopulmonary disease. PCD is also a genetically heterogeneous disorder with many single gene mutations leading to similar clinical phenotypes. Here, we present a novel PCD causal gene, coiled‐coil domain containing 151 (CCDC151), which has been shown to be essential in motile cilia of many animals and other vertebrates but its effects in humans was not observed until currently. We observed a novel nonsense mutation in a homozygous state in the CCDC151 gene (NM_145045.4:c.925G>T:p.[E309*]) in a clinically diagnosed PCD patient from a consanguineous family of Arabic ancestry. The variant was absent in 238 randomly selected individuals indicating that the variant is rare and likely not to be a founder mutation. Our finding also shows that given prior knowledge from model organisms, even a single whole‐exome sequence can be sufficient to discover a novel causal gene.     

Adaptive clustering and adaptive weighting methods to detect disease associated rare variants

Current statistical methods to test association between rare variants and phenotypes are essentially the group-wise methods that collapse or aggregate all variants in a predefined group into a single variant. Comparing with the variant-by-variant methods, the group-wise methods have their advantages. However, two factors may affect the power of these methods. One is that some of the causal variants may be protective. When both risk and protective variants are presented, it will lose power by collapsing or aggregating all variants because the effects of risk and protective variants will counteract each other. The other is that not all variants in the group are causal; rather, a large proportion is believed to be neutral. When a large proportion of variants are neutral, collapsing or aggregating all variants may not be an optimal solution. We propose two alternative methods, adaptive clustering (AC) method and adaptive weighting (AW) method, aiming to test rare variant association in the presence of neutral and/or protective variants. Both of AC and AW are applicable to quantitative traits as well as qualitative traits. Results of extensive simulation studies show that AC and AW have similar power and both of them have clear advantages from power to computational efficiency comparing with existing group-wise methods and existing data-driven methods that allow neutral and protective variants. We recommend AW method because AW method is computationally more efficient than AC method.     

A method to calculate coverage probability from uncertainties in radiotherapy via a statistical shape model

In this paper we describe a technique that may be used to model the geometric uncertainties that accrue during the radiotherapy process. Using data from in-treatment cone beam CT scans, we simultaneously analyse non-uniform observer delineation variability and organ motion together with patient set-up errors via the creation of a point distribution model (PDM). We introduce a novel method of generating a coverage probability matrix, that may be used to determine treatment margins and calculate uncertainties in dose, from this statistical shape model. The technique does not assume rigid body motion and can extrapolate shape variability in a statistically meaningful manner. In order to construct the PDM, we generate corresponding surface points over a set of delineations. Correspondences are established at a set of points in parameter space on spherically parameterized and canonical aligned outlines. The method is demonstrated using rectal delineations from serially acquired in-treatment cone beam CT image volumes of a prostate patient (44 image volumes total), each delineated by a minimum of two observers (maximum six). Two PDMs are constructed, one with set-up errors included and one without. We test the normality assumptions of the PDMs and find the distributions to be Gaussian in nature. The rectal PDM variability is in general agreement with data in the literature. The two resultant coverage probability matrices show differences as expected.     

Classification of missense variants of unknown significance in BRCA1 based on clinical and tumor information

Classification of rare missense variants in disease susceptibility genes as neutral or disease-causing is important for genetic counseling. Different criteria are used to help classify such variants in BRCA1 and BRCA2; however, the strongest evidence tends to come from segregation analysis and observed cooccurrence with known pathogenic mutations, which both require information that is not readily available in most circumstances. A likelihood-based model has been developed, integrating most of the data currently used to classify these variants. We have adapted the original model, including only that information that could be more easily obtained from a cancer genetics laboratory, such as loss of heterozygosity (LOH), grade, and immunohistochemical analysis to assess estrogen receptor (ER) status for the tumors of carriers of these variants. We also considered summary family history (personal or first-degree family history of bilateral breast or ovarian cancer), which was not incorporated into the original model. To test the ability of the modified model to classify missense variants in BRCA1, we analyzed 17 variants, of which 10 have previously been classified as pathogenic mutations or neutral polymorphisms. We also included a prior step consisting of the screening of the variants among 1,000 controls, with which we were able to classify five as neutral, based solely on their observed frequency. We found that combining this relatively easily collected information can be sufficient to classify variants as pathogenic or neutral if tumors from at least three carriers of the same variant can be collected and analyzed.     

Whole-genome sequencing reveals a coding non-pathogenic variant tagging a non-coding pathogenic hexanucleotide repeat expansion in C9orf72 as cause of amyotrophic lateral sclerosis

Motor neuron degeneration in amyotrophic lateral sclerosis (ALS) has a familial cause in 10% of patients. Despite significant advances in the genetics of the disease, many families remain unexplained. We performed whole-genome sequencing in five family members from a pedigree with autosomal-dominant classical ALS. A family-based elimination approach was used to identify novel coding variants segregating with the disease. This list of variants was effectively shortened by genotyping these variants in 2 additional unaffected family members and 1500 unrelated population-specific controls. A novel rare coding variant in SPAG8 on chromosome 9p13.3 segregated with the disease and was not observed in controls. Mutations in SPAG8 were not encountered in 34 other unexplained ALS pedigrees, including 1 with linkage to chromosome 9p13.2-23.3. The shared haplotype containing the SPAG8 variant in this small pedigree was 22.7 Mb and overlapped with the core 9p21 linkage locus for ALS and frontotemporal dementia. Based on differences in coverage depth of known variable tandem repeat regions between affected and non-affected family members, the shared haplotype was found to contain an expanded hexanucleotide (GGGGCC)(n) repeat in C9orf72 in the affected members. Our results demonstrate that rare coding variants identified by whole-genome sequencing can tag a shared haplotype containing a non-coding pathogenic mutation and that changes in coverage depth can be used to reveal tandem repeat expansions. It also confirms (GGGGCC)n repeat expansions in C9orf72 as a cause of familial ALS.     

Performance of computational methods for the evaluation of pericentriolar material 1 missense variants in CAGI‐5

The CAGI‐5 pericentriolar material 1 (PCM1) challenge aimed to predict the effect of 38 transgenic human missense mutations in the PCM1 protein implicated in schizophrenia. Participants were provided with 16 benign variants (negative controls), 10 hypomorphic, and 12 loss of function variants. Six groups participated and were asked to predict the probability of effect and standard deviation associated to each mutation. Here, we present the challenge assessment. Prediction performance was evaluated using different measures to conclude in a final ranking which highlights the strengths and weaknesses of each group. The results show a great variety of predictions where some methods performed significantly better than others. Benign variants played an important role as negative controls, highlighting predictors biased to identify disease phenotypes. The best predictor, Bromberg lab, used a neural‐network‐based method able to discriminate between neutral and non‐neutral single nucleotide polymorphisms. The CAGI‐5 PCM1 challenge allowed us to evaluate the state of the art techniques for interpreting the effect of novel variants for a difficult target protein.     

Accurate prediction of a minimal region around a genetic association signal that contains the causal variant

In recent years, genome-wide association studies have been very successful in identifying loci for complex traits. However, typically these findings involve noncoding and/or intergenic SNPs without a clear functional effect that do not directly point to a gene. Hence, the challenge is to identify the causal variant responsible for the association signal. Typically, the first step is to identify all genetic variation in the locus region, usually by resequencing a large number of case chromosomes. Among all variants, the causal one needs to be identified in further functional studies. Because the experimental follow up can be very laborious, restricting the number of variants to be scrutinized can yield a great advantage. An objective method for choosing the size of the region to be followed up would be highly valuable. Here, we propose a simple method to call the minimal region around a significant association peak that is very likely to contain the causal variant. We model linkage disequilibrium (LD) in cases from the observed single SNP association signals, and predict the location of the causal variant by quantifying how well this relationship fits the data. Simulations showed that our approach identifies genomic regions of on average ∼50 kb with up to 90% probability to contain the causal variant. We apply our method to two genome-wide association data sets and localize both the functional variant REP1 in the α-synuclein gene that conveys susceptibility to Parkinson''s disease and the APOE gene responsible for the association signal in the Alzheimer''s disease data set.     

Mutation of sodium channel SCN3A in a patient with cryptogenic pediatric partial epilepsy

Mutations in the sodium channel genes SCN1A and SCN2A have been identified in monogenic childhood epilepsies, but SCN3A has not previously been investigated as a candidate gene for epilepsy. We screened a consecutive cohort of 18 children with cryptogenic partial epilepsy that was classified as pharmacoresistant because of nonresponse to carbamazepine or oxcarbazepine, antiepileptic drugs that bind sodium channels. The novel coding variant SCN3A-K354Q was identified in one patient and was not present in 295 neurological normal controls. Twelve novel SNPs were also detected. K354Q substitutes glutamine for an evolutionarily conserved lysine residue in the pore domain of SCN3A. Functional analysis of this mutation in the backbone of the closely related gene SCN5A demonstrated an increase in persistent current that is similar in magnitude to epileptogenic mutations of SCN1A and SCN2A. This observation of a potentially pathogenic mutation of SCN3A (Nav1.3) indicates that this gene should be further evaluated for its contribution to childhood epilepsy.     

马凡综合征新发变异一例

目的:对1例不明原因生长发育过快及心脏畸形的患儿进行临床表型及遗传学病因分析。方法:对先证者进行全外显子测序分析,应用Sanger测序技术对全外显子筛出的可疑致病基因的变异位点进行验证,并利用生物信息学软件进行功能预测分析。结果:全外显子测序结果显示,先证者在 FBN1基因的第48外显子上存在c.5846_...     

Tumor characteristics as an analytic tool for classifying genetic variants of uncertain clinical significance

It is important to identify a germline mutation in a patient with an inherited cancer syndrome to allow mutation carriers to be included in cancer surveillance programs, which have been proven to save lives. Many of the mutations identified result in premature termination of translation, and thus in loss-of-function of the encoded mutated protein. However, the significance of a large proportion of the sequence changes reported is unknown. Some of these variants will be associated with a high risk of cancer and have direct clinical consequence. Many criteria can be used to classify variants with unknown significance; most criteria are based on the characteristics of the amino acid change, on segregation data and appearance of the variant, on the presence of the variant in controls, or on functional assays. In inherited cancers, tumor characteristics can also be used to classify variants. It is worthwhile to examine the clinical, morphological and molecular features of a patient, and his or her family, when assessing whether the role of a variant is likely to be neutral or pathogenic. Here we describe the advantages and disadvantages of using the tumor characteristics of patients carrying germline variants of uncertain significance (VUS) in BRCA1, BRCA2, or in one of the mismatch repair (MMR) genes, MLH1, MSH2, or MSH6, to infer pathogenicity.     

Homozygosity mapping on a single patient: identification of homozygous regions of recent common ancestry by using population data

Homozygosity mapping has played an important role in detecting recessive mutations using families of consanguineous marriages. However, detection of regions identical and homozygosity by descent (HBD) when family data are not available, or when relationships are unknown, is still a challenge. Making use of population data from high-density SNP genotyping may allow detection of regions HBD from recent common founders in singleton patients without genealogy information. We report a novel algorithm that detects such regions by estimating the population haplotype frequencies (HF) for an entire homozygous region. We also developed a simulation method to evaluate the probability of HBD and linkage to disease for a homozygous region by examining the best regions in unaffected controls from the host population. The method can be applied to diseases of Mendelian inheritance but can also be extended to complex diseases to detect rare founder mutations that affect a very small number of patients using either multiplex families or sporadic cases. Testing of the method on both real cases (singleton affected) and simulated data demonstrated its superb sensitivity and robustness under genetic heterogeneity.     

一个遗传性对称性色素异常症家系的ADAR基因变异分析CSCD

目的探讨1个遗传性对称性色素异常症(dyschromatosis symmetrica hereditaria, DSH)家系患者的临床特点及基因变异, 明确其致病原因。方法采集先证者及其母亲的外周血样, 应用PCR扩增结合Sanger测序的方法分别对先证者和母亲的ADAR基因进行变异分析, 确定疑似致病变异。同时以100例与本家系无关的正常人作为对照。结果该病例符合DSH的典型表现, 表现为发生在手背, 脚和面部色素沉着、色素减退斑、色素异常斑。Sanger测序显示家系先证者及其母亲均携带ADAR基因第9外显子c.2762+1G>T杂合变异, 100名健康对照均未发现上述变异。根据美国医学遗传学与基因组学学会指南, ADAR基因c.2762+1G>T剪接变异被判定为致病性(PVS1+PM2+PP4)。结论 ADAR基因c.2762+1G>T变异可能为该家系患者的致病原因, 上述结果丰富了ADAR基因的变异谱。