In silico analysis of non-synonymous single nucleotide polymorphisms in human DAZL gene associated with male infertility

ABSTRACT

In living systems the most frequent type of genetic mutation is the single nucleotide polymorphism (SNP). Non-synonymous SNPs (nsSNPs) or missense mutations arise in coding regions of a particular gene. nsSNPs result in a single amino acid substitution which may have effects on the structure and/or function of proteins. Spermatogenesis is a complex process where haploid spermatozoa are formed. The deleted in azoospermia like (DAZL) gene has a relationship with male infertility and dysfunction of DAZL may decrease the sperm count which leads to oligozoospermia or azoospermia. Various computational methods were used to analyze the genetic variations of DAZL affecting the structure and/or function. In the present study, N109T was assigned as the most deleterious or disease related nsSNP by SIFT, MutPred, PolyPhen 2.0, I-Mutant, and MuStab tools. The ConSurf tool showed that functional amino acid residues which are conserved in Human DAZL include the N109T nsSNP. The secondary and tertiary structure was predicted using PSIPRED and MUSTER. Our study shows that the N109T variant may directly or indirectly weaken amino acid interactions and hydrogen bond networks of the DAZL protein, which we predicted may result in altered DAZL protein function. Further, computational analysis of free energy change due to this point mutation using GROMOS96 indicated decreased stability of the DAZL protein. The N109T variant in an infertile male population may provide a genetic marker for mutational analysis of DAZL.

Abbreviations: DAZL: deleted in azoospermia like; dbSNP: database of single nucleotide polymorphism; nsSNPs: non-synonymous SNPs; AA: amino acid; SIFT: sorting intolerant from tolerant; PolyPhen-2: polymorphism phenotyping v2; MUSTER: multi-sources threader; PDB: protein data bank; MuStab: predicting mutant protein stability change; PSIPRED: PSI-blast based secondary structure prediction

KEYWORDS:

• DAZL gene
• in silico analysis
• male infertility
• missense
• nsSNP

Introduction

The DAZL (DAZ-like) gene is a member of DAZ (deleted in azoospermia) family in which Boule (Boll), DAZL, and DAZ genes are included. DAZL is a single copy gene with two isoforms of protein, which is considered to be the ancestor of the DAZ gene [Yang et al. 2005; Kumar et al. 2011; Cauffman et al. 2005]. DAZL maps to the short arm of chromosome 3p24.3 [Yang et al. 2005]. The DAZL protein is expressed in the primordial germ cells (PGCs) and shares about 83% sequence similarity with DAZ [Khabour et al. 2013; Xu et al. 2001]. DAZL codes for RNA binding protein having RNA recognition motif (RRM) and one DAZ repeat which binds to other mRNA binding proteins, controlling the differentiation, growth, and maturation of germ cells [Fu et al. 2015].

Infertility is defined as the failure of a couple to conceive after at least one year of frequent and unprotected intercourse. Its frequency is about 10–15% worldwide, where half likely involves a male factor [Evers 2002; Devroey et al. 2009]. Spermatogenesis is a complicated, multistep process and failure in any steps can lead to oligozoospermia or azoospermia having a substantial genetic component [Foresta et al. 2002]. Approximately 10% of azoospermia and oligospermia patients show a deletion in DAZ family genes which makes DAZL a promising candidate gene for male infertility [Nagafuchi et al. 1993; Reijo et al. 1995; Nakahori et al. 1996; Saxena et al. 1996]. During spermatogenesis, the DAZL protein plays an important role where it might function as a translational activator and present in cytoplasm and nuclei of gonocytes, nuclei of spermatogonia, and cytoplasm during meiosis of spermatocytes [Ruggiu et al. 2000; Seligman and Page 1998]. Lin et al. [2001] reported that the amount of DAZL transcripts in the testes of azoospermic men is lower when compared to fertile men.

In the human genome, numerous SNPs are identified accounting for more than 90% of all sequence variation [Collins et al. 1998]. These SNPs occur in protein coding regions, especially non-synonymous SNPs (nsSNPs) also known as missense variants, where the substitution of amino-acid at the protein level alters the function leading to pathogenic phenotypes [Goswami 2015]. SNPs may or may not affect protein function. Therefore, it is necessary to understand the association between the SNPs and its phenotypic influence which might be useful in analyzing the reasons of numerous diseases or disorders. Several computational methods have been developed which can predict disease associated missense variants. This includes amino-acid substitution that can affect biological functions by altering the structure, folding, or stability of protein [Thusberg et al. 2011].

Currently, computational methods are widely used as a first scan of likely candidates. Many nsSNPs of DAZL are still unclear which may have disease causing potential. Therefore, this study aimed to evaluate the most deleterious nsSNPs in DAZL gene, which may affect the stability and/or function. A set of computational tools that included sorting intolerant from tolerant (SIFT), polymorphism phenotyping v2 (PolyPhen 2), MutPred, I-Mutant 2.0, and MuStab were used to identify candidate deleterious nsSNPs in DAZL. DAZL protein structure prediction/model and conservation analysis were done by ConSurf, PSIPRED, multi-sources threader (MUSTER), project HOPE, and FTsite. The study design is represented schematically in Figure 1.

Figure 1. Diagrammatic representation of computational tools used for in silico analysis of DAZL gene.

Results

In this study, we have used SIFT, MutPred, PolyPhen 2, I-Mutant 2.0, and MuStab computational methods to identify the deleterious nsSNPs mutations in the DAZL gene. Structural, conservation analysis, and secondary structure prediction were done by ConSurf, PSIPRED, and project HOPE. Prediction of DAZL protein 3D structure was done using MUSTER followed by energy minimization using GROMOS96. Ligand binding site was predicted using FTsite tool.

Accumulation of nsSNPs data from dbSNP and the recognition of deleterious nsSNPs using SIFT, PolyPhen 2, and MutPred

NCBI dbSNP is the most extensively used database. A total of 1,591 SNPs (rsIDs) from NCBI dbSNP were analyzed out of which 105 were missense, 46 were coding synonymous, 1,212 were intronic, 99 occurred in the 3′ UTR, 57 occurred in 5′ UTR region, and the rest were other types of SNPs.

Out of 105 missense SNPs subjected to SIFT analysis, only seven have given output. Table 1 shows the seven nsSNPs of which five were predicted to be tolerated (tolerating index (TI) ≥ 0.05) and two were predicted to be damaging, having TI ≤ 0.05. The TI is inversely proportional to the functional effect of amino acid substitution [Kumar et al. 2009]. Amino acid substitution of rs61730117 and rs75931701 were assigned with a score (TI) of 0.

Table 1. Prediction of nsSNPs effect using SIFT.

dbSNP ID	Amino acid substitution	Amino acid	Prediction	Score (Tolerating index)
rs121918346	T74A	T	TOLERATED	1
		A	TOLERATED	0.09
rs11710967	T32A	T	TOLERATED	1
		A	TOLERATED	0.17
rs55900737	S5L	S	TOLERATED	0.95
		L	TOLERATED	0.24
rs61730117	S34C	S	TOLERATED	1
		C	DAMAGING	0
rs73142539	R13C	R	TOLERATED	0.98
		C	TOLERATED	0.09
rs75931701	N109T	N	TOLERATED	1
		T	DAMAGING	0
rs114309315	P19S	P	TOLERATED	0.32
		S	TOLERATED	0.58

nsSNPs: non-synonymous SNPs; SIFT: sorting intolerant from tolerant; dbSNP: database of single nucleotide polymorphism.

Prediction of nsSNPs using PolyPhen 2 revealed three different types of mutation as possibly damaging, probably damaging, and benign involving DAZL, based on false positive rate (FPR) value [Adzhubei et al. 2013]. It was observed that the amino acid substitution S34C and N109T were probably damaging (Table 2), whereas R13C was possibly damaging, and T74A, T32A, S5L, and P19S were benign.

Table 2. List of variants analyzed using PolyPhen 2.

dbSNP ID	Amino acid substitution	HumDiv		HumVar
		Prediction	Score	Prediction	Score
rs61730117	S34C	Probably	0.999	Possibly	0.906
rs75931701	N109T	Probably	0.974	Possibly	0.665

PolyPhen-2: polymorphism phenotyping v2; dbSNP: database of single nucleotide polymorphism.

A missense mutation having a MutPred score > 0.5 could be considered as “harmful” and a score > 0.75 should be treated as a high confidence “harmful” prediction [Li et al. 2009]. Among the seven nsSNPs, one was found to be a harmful mutation with a score of > 0.5 and the remaining six nsSNPs were found to be normal with the score < 0.5. The probability of a deleterious mutation score for N109T was 0.580 and loss of stability was predicted (p = 0.0027). However, S34C did not lead to loss of stability in DAZL (Table 3).

Table 3. Prediction of nsSNPs effect in DAZL structure, function, and evolution using MutPred.

dbSNP ID	Mutation	Probability of deleterious mutation	Molecular mechanism disrupted	Top 5 features
rs61730117	S34C	0.251		Loss of glycosylation at S34 (p = 0.0046) Loss of phosphorylation at S34 (p = 0.0123) Gain of catalytic residue at S38 (p = 0.0515) Loss of loop (p = 0.2897) Gain of sheet (p = 0.3763)
rs75931701	N109T	0.580	Actionable Hypotheses: Loss of stability (P = 0.0027)	Loss of stability (p = 0.0027) Loss of helix (p = 0.1299) Gain of sheet (p = 0.1451) Gain of catalytic residue at N109 (p = 0.3052) Gain of phosphorylation at S106 (p = 0.565)

nsSNPs: non-synonymous SNPs; DAZL: deleted in azoospermia like; dbSNP: database of single nucleotide polymorphism.

Analysis of stability changes due to nsSNPs found by I-Mutant 2.0 and MuStab

The stability of seven nsSNPs associated with DAZL protein was predicted by I-Mutant 2.0 and MuSTAB through comparing free energy. I-Mutant predicts the stability of the protein upon amino acid substitution by examining the Gibbs free energy by ∆∆G value = ∆G (New protein) - ∆G (Wild type) in kcal/mol, which is calculated at pH 7 and 25°C. The DDG prediction by I-Mutant 2.0 showed that the five nsSNPs (rs121918346, rs11710967, rs73142539, rs75931701, and rs114309315) had decreased stability value with DDG<0 whereas two nsSNPs (rs55900737 and rs61730117) had an increased stability value with DDG>0. According to MuSTAB, four nsSNPs (rs121918346, rs73142539, rs75931701, and rs114309315) showed decreased stability and three nsSNPs (rs11710967, rs55900737, and rs61730117) showed increased stability. Two nsSNPs (rs61730117 and rs75931701) which were selected through SIFT and PolyPhen2 and their comparative analysis using I-Mutant 2.0 and MuStab showed that only one nsSNP (rs75931701) had a decreased stability (DDG value -1.09 and prediction confidence 82.32%) that would be predicted to disturb the structure and function of the protein (Table 4).

Table 4. Stabilities of nsSNPs by I-Mutant 2.0 and MuSTAB.

dbSNP ID	Amino acid substitution	I-Mutant 2.0			MuSTAB
		Stability	RI	DDG	Protein stability	Prediction confidence
rs61730117	S34C	Increase	1	-0.07	Increase	25.89%
rs75931701	N109T	Decrease	1	-1.09	Decrease	82.32%

nsSNPs: non-synonymous SNPs; dbSNP: database of single nucleotide polymorphism; MuStab: predicting mutant protein stability change.

Structural conformation and conservation analysis by ConSurf and secondary structure prediction by PSIPRED

Figure 2 represents the result predicted by the ConSurf tool which contains nine color codes based on conservation score, which indicate evolutionary relationships among their sequence homologs [Celniker et al. 2013]. It was noticed that variants R13C and P19S had a conservation scale of one; T74A and T32A had a conservation scale of two; S5L had a conservation scale of three; and S34C and N109T had a conservation scale of six.

Figure 2. Analysis of evolutionary conserved amino acid residues of DAZL by ConSurf. The color coding bar shows conservation score.

Secondary structure of DAZL was predicted by PSIPRED which explained the distribution pattern of alpha helix, beta sheet, and coil (Figure 3). The result indicated higher percentage of coils (84.76%) followed by beta sheet (9.21%) and alpha helix (6.03%) in the predicted secondary structure.

Figure 3. Secondary structure of DAZL gene showing alpha sheet, beta helix, and coil.

Three-dimensional protein structure prediction by MUSTER

A total of 10 3D models were provided with Z-score generated by MUSTER for the human DAZL protein. The model which had a Z-score greater than 7.5 is considered as a good template [Wu and Zhang 2008]. The 3D DAZL model selected has a Z-score 8.809 for wild type, and for mutant (N109T) the Z-score is 8.199. The 3D model selected for both wild and mutant is considered as good by MUSTER. Both 3D models were analyzed using RasWin Molecular Graphics. The position of amino acid (AA) alteration at 109 position within DAZL wild type (ASN109) and mutant type (THR109) protein was labelled (Figure 4). Further, structural change of AA in native and mutant protein chains is shown in Figure 5 using RasWin Molecular Graphics. The energy minimization of both wild and mutant structures was done by GROMOS96. The energy minimization value of mutant model is high as compared to the native model (Table 5). Superimposed structure of the protein in ribbon-presentation, showing protein colored in grey, and side chains of both the wild-type and the mutant residue colored in green and red, respectively, for N109T variant by project HOPE (Figure 6).

Table 5. Energy minimization of DAZL 3D protein structure.

Model	Electrostatic	Energy minimization
Wild	-10568.59	-10355.229
N109T	-10224.14	-9844.786

Figure 4. The position of amino acid alteration at 109 position within DAZL wild type (ASN109) and mutant type (THR109) protein. (A) Wild-type model showing asparagine at position 109 (ASN109) of DAZL protein constructed using multi-sources threader (MUSTER) and visualized by RasWin Molecular Graphics. (B) Mutant model showing threonine at position 109 (THR109) of DAZL protein constructed using MUSTER and visualized by RasWin Molecular Graphics.

Figure 5. Structural alteration in DAZL at amino acid position 109. (A) Due to asparagine and (B) due to threonine.

Figure 6. Superimposed structural alteration at 109 Asn (A) and 109 Thr (B) residues within DAZL protein.

Ligand binding site prediction by FTsite

The FTsite algorithm identifies binding sites using apo structures from two established test sets. The FTsite tool identified three different ligand binding sites of the DAZL (Table 6). The ligand binding site 1 consists of 14 residues, site 2 consists of 11 residues, and site 3 consists of 8 residues. The N109T was predicted as deleterious by various computational tools also positioned at first and third ligand binding site (ASN-109) confirming its strong deleterious effect on DAZL protein functioning (Figure 7).

Table 6. Residues at ligand binding sites of DAZL.

Site 1	Site 2	Site 3
TYR X 83	LYS X 56	ASN X 109
GLY X 84	ILE X 57	VAL X 111
SER X 85	MET X 58	ASN X 165
PHE X 107	PRO X 59	PRO X 166
PHE X 108	PRO X 132	ASN X 167
ASN X 109	ALA X 133	THR X 168
VAL X 111	ILE X 134	TYR X 171
ASP X 112	ARG X 135	TYR X 202
LYS X 115	ARG X 221
ASN X 165	VAL X 232
GLN X 197	ASN X 233
THR X 200
TYR X 202
VAL X 246

Figure 7. Ligand binding site prediction using FTSite showing ASN-109 residue at site 1 and 3. Mesh (A: light gray (salmon) and B: darker gray (blue) colored) representations of the sites and sticks of the residues around the binding sites within DAZL protein.

Discussion

There are many genes involved in male infertility and defects or loss of these genes may result in infertility. There is very finite knowledge about the genetic variations which affect the expression level and also the phenotypic variation of spermatogenesis. DAZL is one of the important candidates for autosomal recessive infertility. The DAZL gene is a member of the deleted in azoospermia (DAZ) gene family, which was previously symbolized as DAZLA (DAZ-Like-Autosomal). Both DAZ and DAZL genes are functionally complementary to one another. During primate evolution the DAZ gene arose by the transposition and amplification of the DAZL gene from chromosome 3 to Yq [Saxena et al. 1996; Bartoloni et al. 2004]. DAZL is a RNA binding protein that is expressed both in male and female germ cells. Hence, it is an essential gene, which plays a crucial role in reproduction. In the testis, the expression of the DAZL transcript is higher. DAZL is expressed in primary spermatocytes and spermatogonia and males with spermatogenic defects have a lower level of DAZL transcript in the testis [Lin et al. 2001].

Single nucleotide polymorphism (SNP) data are available in the human genome and is expanding, providing pathways for future research. Genetic variations like non-synonymous single nucleotide polymorphisms (nsSNPs) are the most familiar type of variation in humans where AA substitution in the protein occurs, also known as missense mutations. Missense mutations prevail more than other variants, which account for almost half of all the allelic variants which fall in inherited human diseases [Hamosh et al. 2005]. In silico analysis is now used to predict disease related SNPs at the molecular level. In this study, we predicted the pathogenic missense mutations located in the DAZL gene, which can be associated with male infertility. It is useful to study the effect of missense mutations to give insight into the molecular basis of hereditary diseases. Therefore, an effort was made to predict nsSNPs that can change the structure, function, and expression level of the DAZL gene. Out of seven nsSNPs, two were found to be deleterious by SIFT, one by MutPred, all of them by PolyPhen 2.0, five by I-Mutant, and four by MuStab. One crucial point mutation N109T (rs75931701) in the coding region was predicted to be deleterious by all the computational methods which may have a significant impact on DAZL structure and/or function. This nsSNP was not reported in any previous study which shows association with male infertility. Hence, it is necessary to validate this nsSNP to support this finding. In silico analysis can help to evaluate the probability that a nsSNP is deleterious for DAZL function [Squitti et al. 2014] by comparing the extent to which an SNP will influence the stability of the mutated protein with respect to the native type. I-Mutant and MuSTAB predicted that the N109T variant of DAZL could affect the stability of the folded protein [Capriotti et al. 2005].

Functional regions based on evolutionary rate was predicted using ConSurf. The evolutionary rate was calculated depending on evolutionary relatedness among proteins and homologos [Ashkenazy et al. 2010] for predicting a mutation that may have an affect [Ramensky et al. 2002]. N109T showed the conservation scale of 7 which indicates the mutation is slowly evolving.

Based on the tertiary structure of a protein, it interacts with other biomolecules or imparts different functions. Thus, it is necessary to predict the tertiary structure of the DAZL gene as there is no crystal structure of DAZL available in the Protein Data Bank (PDB). The3D structure of DAZL has been modeled using MUSTER. Energy minimization of both the native type protein (DAZL) and the mutant type protein model revealed that the total energy of the mutant protein structure was different from the native type protein structure. The total energy of native and mutant models after energy minimization was -10355.229 KJ/mol and -9844.786 KJ/mol, respectively. The total energy of the mutant model is higher as compared to the native model which suggests that the mutation decreases the stability of the DAZL protein. Project HOPE 3D protein structure for N109T suggests that the mutant residue is smaller than the wild-type residue which causes a possible loss of external interactions. Additionally, there is a difference between the hydrophobicity of the wild and mutant DAZL. The wild-type residue is more hydrophobic than the mutant residue. The FT site method is an accurate method to identify binding sites of nsSNPs as it is based on experimental evidence. Binding site evaluation is important as it is applied to structure-based prediction of protein function, knowing functional links between proteins, protein engineering, and drug design. N109 is predicted as a binding site which may thus alter the ligand binding affinity of the DAZL protein [Jawon and Kong 2004]. Our study suggests that this disease related SNP should be considered as an important candidate in DAZL dysfunction which could be helpful in further research of genetically inherited disease. The proposed mutant predicted deleterious can be used for drug discovery, pharmacogenomics, and pharmacokinetic studies. The 3D structure will provide a good platform for functional analysis of experimentally acquired crystal structures. Therefore, it may be possible that SNPs in the protein can affect its interaction with other molecules or part of the protein. The N109T variant can serve as a genetic marker that may be used for mutational analysis of the DAZL gene for male infertility.

Materials and methods

Datasets for DAZL (Homo sapiens) gene

Data on the human DAZL gene were gathered from the National Centre for Biotechnology Information- NCBI, (http://www.ncbi.nlm.nih.gov/) and Online Mendelian Inheritance in Man-OMIM, (http://www.omim.org/entry/601486). SNP data of the DAZL gene were obtained from the National Centre for Biological Information (NCBI) dbSNP (http://www.ncbi.nlm.nih.gov/snp/) and each SNP has their reference sequence ID (rsID). Sequence of the DAZL gene was obtained from NCBI (http://www.ncbi.nlm.nih.gov/gene/1618) and the AA sequence of the DAZL gene was obtained from UniProt database (http://www.uniprot.org/uniprot/Q92904). All this information is used for further computational analysis.

Prediction methods

Prediction of tolerated and deleterious SNPs by SIFT

The sequence homology SIFT algorithm, a web based tool (http://sift.jcvi.org) to predict the effect of a substitution based on AA change, was used. Those AA positions in a protein sequence which are conserved throughout evolution tend to be intolerant to the substitution and may affect the protein function. The ability of SIFT to differentiate between neutral and deleterious nsSNPs favors the use of SIFT as a prediction tool [Kumar et al. 2009]. The rsIDs of the DAZL gene from NCBI were entered to collect the information .

Prediction of functional alteration of coding nsSNPs by PolyPhen-2

PolyPhen 2 is an online tool which investigates the AA substitution effect on the structure and function of human proteins. Prediction by PolyPhen 2 is based on the number of sequences, phylogenetic, and structural properties characterizing the substitution. It is important to have automated predictions for interpreting the genetic variants from large datasets which are helpful in human genetic research. The PolyPhen 2 web interface can be achieved at http://genetics.bwh.harvard.edu/pph2/ [Adzhubei et al. 2013].

We entered a protein identifier (UniProtKB accession number or entry name) or protein sequence in FASTA format of the DAZL gene from NCBI (NP_001177740.1) followed by entering the position of the substitution in the AA sequence. Selecting the appropriate boxes for the wild-type AA residue in AA1 and the substitution residue in AA2.

Prediction of harmful mutations by MutPred

The MutPred tool was developed to predict the single site mutation as disease related or neutral in human based on protein structure, function, and evolution. MutPred is a Random Forest (RF) based method provide “g” score for deleterious substitution and “p” score for molecular mechanism distribution [Li et al. 2009]. This tool is freely available at http://mutpred.mutdb.org/. It was used to introduce the sequence of the protein of DAZL gene in FASTA format from NCBI (NP_001177740.1) and the mutations (nsSNPs) to predict the harmful mutation.

Calculation of stability of predicted nsSNPs by free energy

The nsSNPs mostly alter the structural stability of the protein, which affects protein function. Therefore, it is one of the necessary parameters to check the stability of deleterious nsSNPs. In order to predict the stability, we used two different web servers- I-Mutant 2.0 and MuStab [Raghav and Sharma 2013].

I-Mutant 2.0

I-Mutant 2.0 server is available at http://folding.biofold.org/i-mutant/i-mutant2.0.html. It is a support vector machine (SVM) based tool for online automatic prediction of stability change in protein due to nsSNPs. This tool is used for predicting the sign of protein stability changes and ∆∆G association values upon mutation [Capriotti et al. 2005; Raghav and Sharma 2013].

To get the information related to protein stability, the protein sequence from the UniProt database (Q92904) was taken followed by substitution position and new AA after single site mutation.

MuSTAB

MuSTAB (http://bioinfo.ggc.org/mustab/) is another tool which is also based on SVM, to check the protein stability changes upon single-site mutations [Teng et al. 2010; Raghav and Sharma 2013]. We entered the AA sequence of human DAZL gene in FASTA format from NCBI (NP_001177740.1) and defined the position and the identity of the substituting residue to collect the output data.

Structural and conservation analysis of DAZL

ConSurf server is a freely available tool which predicts the evolutionary conservation based on phylogenetic relationship among homologous sequences of AA positions in protein or nucleic acid positions in DNA/RNA. This tool predicts the conservation score based on sites evolutionary rate using empirical Bayesian [Mayrose et al. 2004] or maximum likelihood [Martz 2002] paradigm. The protein sequence in FASTA format from NCBI (NM_001190811.1) was pasted to know the evolutionary conservation of AAs within the protein.

Secondary structure prediction of DAZL by PSIPRED

This tool is used to predict the protein secondary structure based on their position specific matrices which is developed by PSI-BLAST [Altschul et al. 1997]. PSIPRED prediction server is available at http://bioinf.cs.ucl.ac.uk/psipred/. Secondary structure of DAZL was predicted by entering single or multiple sequence alignments in raw sequence or in FASTA format.

MUSTER-3D structure predictor

This threading algorithm MUSTER tool, combines different sequences and structure data which can be used in dynamic programming search: (i) sequence-derived profiles, (ii) secondary structures, (iii) structured-derived profiles, (iv) solvent accessibility, (v) backbone torsion angles, and (vi) hydrophobic scoring matrix [Wu and Zhang 2008]. The MUSTER tool is freely available at http://zhanglab.ccmb.med.umich.edu/MUSTER/. It predicts the structure of protein by using MODELLER v8.2 and calculating Z-score to classify the protein sequence as good (Z-score>7.5) or bad (Z-score<7.5). The pasted sequence in FASTA format followed by entering e-mail and the ID (optional) as an input query was used to collect prediction data. Further extension was carried out using the HOPE project which provides 3D structural visualization of mutated proteins using UniprotKB and DAS-servers [Venselaar et al. 2010].

FTsite server-identification of nsSNPs on ligand binding sites

FT-site server is an energy based method, which identifies the binding sites accurately over 94% of the apo proteins from two test sets which have been used to identify other binding site prediction methods [Ngan et al. 2012]. PDB file of native protein structure of DAZL was uploaded to predict the binding site.

Declaration of interest

The authors report no declarations of interest.

Acknowledgments

The authors express appreciation to Charutar Vidya Mandal (CVM) and SICART, Vallabh Vidyanagar, Gujarat for providing research work platform. We acknowledge Director, Ashok and Rita Patel Institute of Integrated Study and Research in Biotechnology and Allied Sciences (ARIBAS), New Vallabh Vidynagar for all the facilities and constant encouragement to carry out this work. This work was supported by INSPIRE division of the Department of Science and Technology in the form of a research fellowship.

Additional information

Notes on contributors

Mili Nailwal

Performed methodology, designed the study, analyzed data, and wrote the manuscript: MN; Contributed to study design, interpretation of data, and manuscript revision: JBC.