Research Article, J Blood Res Hematol Dis Vol: 2 Issue: 1

Haemophilia B PMS1 Gene SNPs Predictions through Insilico Approach at Proteomics Levels

Hiba Siddig Ibrahim^*

The National Ribat University, Department of Microbiology, Faculty of Medical Laboratory Science, The National Ribat University, Sudan

*Corresponding Author : Hiba Siddig Ibrahim Mustafa
Department of Microbiology, Faculty of Medical Laboratory Science, The National Ribat University, Sudan
Tel: 249155772782
E-mail: hibasiddig55@gmail.com

Received: December 22, 2016 Accepted: January 05, 2017 Published: January 11, 2017

Citation: Ibrahim HS (2017) Haemophilia B PMS1 Gene SNPs Predictions through In-silico Approach at Proteomics Levels. J Blood Res Hematol Dis 2:1.

Abstract

Keywords: Haemophilia B; Clotting factors; Meta-snp prediction

Introduction

Hemophilia B is characterized by deficiency in factor IX clotting activity due to mutation in factor IX gene, that results in prolonged oozing after injuries, tooth extractions, or surgery, and delayed or recurrent bleeding prior to complete wound healing. Hemophilia B is phenotypically indistinguishable from hemophilia A, it was first recognized as a different kind of hemophilia in 1952 with additional name Christmas disease; named after Stephen Christmas, the first patient who described with this disease. According to the level of factor IX clotting activity beside age of patients and frequency of bleeding episodes hemophilia are divided to 3 levels; mild, moderate and severe [1,2].

Individuals with mild hemophilia B do not have spontaneous bleeding and they often diagnosed later in their life with bleeding tendency once a year or once every ten years; however, without pre- and post-operative treatment, abnormal bleeding can happens with surgery or tooth extractions, while in moderate hemophilia B they have spontaneous bleeding with prolonged or delayed oozing after relatively minor trauma and are usually diagnosed before age five to six years; the frequency of bleeding episodes varies from once a month to once a year, but in severe hemophilia B, spontaneous joint or deep-muscle bleeding is the most frequent symptom, those individuals are usually diagnosed during the first two years of life; without prophylactic treatment, they may average up to two to five spontaneous bleeding episodes each month [1].

Genetically; factor IX gene is located on the X chromosome (Xq27.1-q27.2), it’s an X-linked recessive trait, which explain why only males usually affected. In addition in 1990, George Brownlee and Merlin Crossley showed that two sets of genetic mutations were preventing two key proteins from attaching to the DNA of people with a rare and unusual form of haemophilia B and haemophilia B Leyden, those people are suffering from excessive bleeding in childhood but it became rare after puberty, This deficiency lead to turning off the gene that produces clotting factor IX, which prevents excessive bleeding [2,3].

In this study I used different computational methods to identify the PMS1 (hemophilia B) gene SNPs and to predict mutation effects at the proteomic level.

Materials and Methods

Hemophilia B retrieved sequence

Hemophilia B sequence was retrieved from NCBI through SNP search (https://www.ncbi.nlm.nih.gov/snp/?term=haemeophilia%20B); rs1145234 [Homo sapiens] in chromosome two (Chromosome: 2:189867833) for PMS1 gene, it is intr on variant, missense, NC transcript variant Allele Origin: G (germline)/T (germline)/C (germline), this was selected from total SNPs 2347417 results.

Proven

Proven protein batch for human (PROVEAN v1.1.3.) (http://provean.jcvi.org/index.php) was selected in this study; PROVEAN is abbreviation for Protein Variation Effect Analyzer, its useful software for filtering sequence variants to identify nonsynonymous or indel variants that are predicted to be functionally important [4-6].

Sift Prediction (SIFT - Predict effects of nonsynonmous / missense variants)

(http://sift.bii.a-star.edu.sg/) SIFT dbSNP 138 was selected from batch tools from SIFT Sorting Intolerant From Tolerant software to predict whether an amino acid substitution affects protein function based on sequence homology and the physical properties of amino acids.

PolyPhen-2 (Polymorphism phenotyping v2)

PolyPhen-2 prediction of functional effects of human nsSNPs (http://genetics.bwh.harvard.edu/pph2/index.shtml) was used to predict the impact of an amino acid substitution on the structure and function of a human protein using straightforward physical and comparative considerations [7-10].

I-Mutant suite

I mutant.3 (http://gpcr2.biocomp.unibo.it/cgi/predictors/IMutant3.0/I-Mutant3.0.cgi) was used to predict protein stability changes upon single point mutation from: Protein Structure / Protein Sequence and also prediction of disease associated with single point mutation from Protein Sequence [11-13].

SNPs and GO

SNPs & GO (http://snps.biofold.org/snps-and-go/snps-and-go.html) was used to predicting a disease associated variations by GO terms through SVM-based classifier, which takes in input protein sequence, profile and functional information to give output as disease-related or neutral with RI & scores [14-19].

Meta-SNPMeta-SNP (Meta-predictor of disease causing variants) (http://snps.biofold.org/meta-snp/#) was used to confirm the results of previous methods; PANTHER, PhD-SNP, SIFT and SNAP, through the non-synonymous SNVs (nsSNVs), that’s results in a single amino acid substitutions and may affect protein functions that lead to causing a disease [20-22].

Results

Proven

The total of 143 nsSNPs which was predicted by Proven prediction tools showed 83 snSNPs which are deleterious with in these 81 nsSNPs were damaging and 3 nsSNPs were tolerated according to sift prediction while 45 nsNSPs showed neutral prediction by proven with 33nsSNPs were damaging and 12 were tolerated according to Sift prediction but 15 snSNPs they were considered benign by polyphen-2 which were excluded from this study, please see details below with Table 1.

In rs1145232; proven prediction showed 5 nsSNPs were deleterious with sift prediction damaging for all.

In rs5742973, according to proven prediction, 1 nsSNPs showed deleterious with sift prediction damaging.

In rs56309301; 1 nsSNPs showed neutral prediction by proven while sift prediction showed tolerated.

In rs61756360; according to proven prediction, 1 nsSNPs showed deleterious with sift prediction showed damaging, in rs111254723; 4 nsSNPs showed deleterious according to proven with sift prediction tolerated for all of them, in rs142159998; 3 nsSNPs showed deleterious by proven while sift prediction showed damaging for all, in rs143010673; according to proven prediction, 3 nsSNPs showed deleterious with sift prediction damaging for all, in rs143554211; 6 nsSNPs showed neutrals according to proven prediction with sift prediction showed damaging for all, in rs143686298; proven prediction showed 2 nsSNPs with neutrals prediction and damaging prediction by sift for all, in rs145103030; 5 nsSNPs showed deleterious according to proven prediction and damaging by sift prediction, in rs145521752; 6 nsSNPs where deleterious by proven prediction while sift prediction showed damaging for all, in rs147901996; 3 nsSNPs they were deleterious by proven with damaging sift prediction.

In rs148745528; 5 nsSNPs showed deleterious prediction by proven with damaging sift prediction predicts, in rs185306467; 3 nsSNPs showed deleterious prediction by proven with sift prediction damaging for the three, in rs188947672; 4 nsSNPs showed neutrals proven prediction and sift prediction damaging for all, in rs189785572; proven prediction showed 2 nsSNPs with neutral prediction and damaging sift prediction for both, in rs193252599; proven predictions showed 2 nsSNPs predict neutral with sift prediction tolerated and damaging for each one of them, in rs199892342; proven prediction showed 1 nsSNPs with neutral prediction while sift prediction showed damaging predict, in rs200172149; according to proven prediction 1 nsSNPs showed neutral predict with tolerated sift prediction.

In rs200244068; proven prediction showed 2 nsSNPs deleterious in addition to damaging sift prediction for both, in rs200395679; proven prediction showed 1 nsSNPs showed neutral with damaging sift prediction, in rs200919195; according to proven prediction 1 nsSNPs showed neutral prediction with tolerated sift prediction, in rs201944922; 1 nsSNPs showed deleterious according to proven prediction while sift prediction showed damaging effect, in rs368282144; 5 nsSNPs showed deleterious proven prediction while sift prediction showed 3 nsSNPs with damaging and 2 nsSNPs they were tolerated.

rs369226504; according to proven prediction, 3 nsSNPs showed deleterious with 3 nsSNPs neutral while for sift prediction all of them showed damaging, in rs369814333; in proven prediction, 1 nsSNPs showed deleterious and the second one showed neutral with tolerated sift prediction for both, rs369909997; by proven prediction showed 1 nsSNPs neutral and tolerated sift prediction, rs370084230; 1 nsSNPs showed deleterious by proven prediction and damaging sift prediction, in rs370300640; According to proven prediction, 1 nsSNPs showed deleterious with sift prediction damaging, in rs370578519; 4 nsSNPs showed deleterious by proven prediction with damaging sift prediction for all, in rs370668897; 1 nsSNPs showed deleterious according to proven prediction and damaging sift prediction, in rs371745827; 6 nsSNPs showed deleterious by proven prediction and all of them were damaging by sift prediction, in rs371999153; 1nsSNPs showed deleterious by proven prediction and damaging by sift prediction, in rs372752293; 2 nsSNPs showed deleterious according to proven and damaging by sift prediction, in rs374222815; 1 nsSNPs showed neutral by proven prediction and damaging according to sift prediction, in rs374879205; 5 nsSNPs showed deleterious according to proven prediction, with damaging sift prediction for all five, in rs375020232, proven prediction showed 6 nsSNPs with deleterious predict and damaging predicts by sift prediction, in rs375164425; 2 nsSNPs showed neutral predict by proven and tolerated predicts by sift prediction.

In rs375553851; 4 nsSNPs showed neutral predicts by proven prediction but 1 nsSNPs deleterious predict while the forth were had damaging predicts by sift prediction, another rs375640863; showed 4 nsSNPs with deleterious predicts by proven predict and damaging by sift prediction, in rs377024581; according to proven prediction 1nsSNPs showed deleterious predict with damaging sift prediction, rs377266054; proven prediction showed 2 nsSNPs with neutral predicts in addition to damaging sift prediction and the last one rs377603311; showed 4 nsSNPs with deleterious predicts by proven prediction and 1 nsSNPS with neutral predict while sift prediction showed damaging for all (Table 1).

Table 1: Illustrate proven prediction with sift results for SNPs.

Sift prediction

A total numbers of 127 out of 143 nsSNPs were considered deleterious by sift prediction in addition to 15 other snSNPs that was also deleterious too but they excluded from this study due to they considered benign according to polyphen-2 software with 1 snSNPs in rs188947672 at position L155V which represented the only SNPs that showed tolerated prediction by sift (Table 2).

Table 2: Illustrate SNPs predictions results by SIFT and polyphen-2 prediction tools.

Polyphen-2

The Polyphen-2 results showed 85 nsSNPs out of 143 nsSNPs as probably damaging, while 27 nsSNPs showed possibly damaging with16 nsSNPs that showed no results; they are (rs1145232/ rs111254723/ rs143010673/ rs143554211/ rs145103030/ rs145521752/ rs147901996 / rs148745528 / rs185306467/ rs188947672/ rs189785572/ rs193252599/ rs199892342/ rs368282144/ rs369226504/ rs369814333/ rs370578519/ rs371745827/ rs374222815/ rs374879205/ rs375020232/ rs375164425/ rs375553851/ rs375640863/ rs377266054/ rs377603311 at the following positions: G325R, Y751C, R721C, W22C, R695C, R743C, E676A, N145S, K138E, I21N, L591V, A386S, E604K, V671A, P338A, T638R, N517I, P524L, R636C, V706M, P713A, R636L, I96V, N588H, D320G, E511V, R707H, with ENSP00000404492 protein ID ); may be due to some sequence problems or even the software and 15 nsSNPs that were considered benign which was excluded from this study (Table 2).

I Mutant suit

According to I mutant suite-3 results I found 103 snSNPs showed decreased in protein functionality within these 103 snSNPs; 14 snSNPs were considered benign by polyphen-2 and should be excluded from this study in addition 14 snSNPs showed increased in protein functionality while 26 snSNPs were not showed any results (Table 3).

Table 3: Illustrate Imutant suite 3 results.

SNPs and GO

According to SNPs & GO prediction I showed the following results:-

PhD-SNP prediction results showed 70 snSNPs predicted as disease in addition to other 9 nSNPs that were excluded due to they gave benign by Polyphen-2, 32 nSNPs considered neutral in addition to other 6 snSNPs which was considered benign according to Polyphen-2 and excluded from this study too while 26 snSNPs that showed no prediction results.

SNPS & GO prediction showed 53 snSNPs that predicted as disease and 49 snSNPs as neutral in addition to 15 snSNPs that considered benign by Polyphen-2 while 26 snSNPs that did not showed any prediction results; both benign and non-prediction snSNPs they are excluded from this study.

Meta-SNP

According to Meta-SNP prediction I showed the following results:-

PhD-SNP showed 69 snSNPs considered as disease in addition to other 9 nSNPs that were excluded due to they gave benign by Polyphen-2, 32 snSNPs were considered neutral in addition to other 6 nSNPs that were excluded due to they gave benign by Polyphen-2 while 26 snSNPs showed no predictions result, both benign and nonprediction snSNPs they are excluded from this study.

SIFT results showed 78 snSNPs as disease in addition to other 10 snSNPs that were excluded due to they predict benign by Polyphen-2, 22 snSNPs showed neutral in addition to other 5 snSNPs that where considered benign by Polyphen-2, 2 snSNPs showed NA and 26 snSNPs showed no results both benign and non-prediction snSNPs they are excluded from this study.

SNAP prediction illustrate 76 snSNPs as disease in addition to other 8snSNPs that excluded considered by Polyphen-2 due to benign predictions, 26 snSNPs they are neutral with 7snSNPs as benign by polyphen-2 and 26 snSNPs without any information.

Meta- SNP prediction showed 68 snSNPs as disease disease in addition to other 7 snSNPs that considered benign by polyphen-2 and exclude from this study; 34 snSNPs predicted as neutral disease in addition to other 8 snSNPs which was considered benign by – polyphen-2 and 26 snSNPs without any prediction information, those without prediction with those considered benign by polyphen-2 are excluded from this study too (Table 4).

Table 4: Illustrate SNPs & GO, Meta-SNP prediction results.

Discussion

Hemophilia is one of the hematologic important disease especially type A then followed by type B and C here in Sudan we showed hemophilia in a higher numbers of population may be due to habits of relative marriage beside cost of treatment was very expensive; in this study I used in-silico methods in translation topic to detect some SNP effects on already available untested database for hemophilia B; according to those tools that I was already used such as Proven, sift, polyphen-2, I mutant suite 3, SNPs & GO and Meta- SNP prediction, I found the previous results in addition to that there are some differences between these methods results specially when trying to confirm their results with another tools, which represented in the followings; In rs56309301→N243T it represented deleterious, tolerated, neutral, neutral according to sift, sift & proven, sift in metasnp sequential, in rs111254723→(Y550C, Y888C, Y927C) represented deleterious, neutral, tolerated, disease; rs111254723→(Y751C) represented deleterious, neutral, tolerated, without sift from metasnp prediction; rs143010673→(W137C) deleterious, deleterious, damaging, neutral; rs143554211→(R695C) deleterious, neutral, damaging, without prediction; rs143554211→(R494C) deleterious, neutral, damaging, neutral; rs143554211→(R259C, R832C, R871C, R709C) deleterious, neutral, damaging, disease; rs143686298 →(P732T, P771T) deleterious, neutral, damaging, neutral; rs147901996→(N106S) deleterious, deleterious, damaging, neutral; rs148745528→(K99E) deleterious, deleterious, damaging, NA; rs148745528→(K253E) deleterious, deleterious, damaging, neutral; rs185306467→(I136N, I197N) deleterious, deleterious, damaging, neutral; rs188947672→(L706V, L728V) deleterious, neutral, damaging, disease; rs188947672→(L155V) tolerated, neutral, damaging, disease; rs189785572→( A501S, A386S) deleterious, neutral, damaging, disease/ without meta-snp prediction; rs193252599→(E168K, E604K) deleterious, neutral, tolerated/damaging, disease/ without meta-snp prediction; rs199892342→( V671A) deleterious, neutral, damaging, without meta-snp prediction; rs200172149→(T202A) deleterious, neutral, tolerated, neutral; rs200244068→(D106G, D167G) deleterious, deleterious, damaging, neutral; rs200395679→(A104T) deleterious, neutral, damaging, disease; rs200919195→(Y195F) deleterious, neutral, tolerated, neutral; rs368282144→( P299A, P453A, P338A) deleterious, deleterious, damaging, neutral/ neutral/without meta-snp prediction; rs368282144→(P514A, P475A) deleterious, deleterious, tolerated, disease; rs369226504→(T202R, T775R, T814R) deleterious, deleterious, damaging, disease; rs369226504→(T437R, T652R, T638R) deleterious, neutral, damaging, disease/ disease /without meta-snp prediction; rs369814333→(N81I, N517I) deleterious, deleterious/neutral, tolerated, neutral/ without meta-snp prediction; rs369909997→(N228D )deleterious, neutral, tolerated, neutral; rs370084230→(R10Q) deleterious, deleterious, damaging, neutral; rs370668897→(K51T )deleterious, deleterious, damaging, neutral; rs372752293→(T49R) deleterious, deleterious, damaging, neutral; rs374222815 →(V505M, V843M, V882M, V720M, V706M) deleterious, neutral, damaging, disease/ disease/ disease/ disease/ without meta-snp prediction ; rs374879205→(P512A, P713A, P850A, P889A, P727A) deleterious, deleterious, damaging, neutral /without meta-snp prediction /disease/disease/disease; rs375020232→(R200L) deleterious, deleterious, damaging, disease; rs375164425→(I57V, I96V) deleterious, neutral, tolerated, NA/ without meta-snp prediction; rs375553851→(N703H, N152H, N588H, N725H, N764H) deleterious, neutral/ deleterious/neutral/ neutral/ neutral, damaging, disease; rs375640863→(D435G, D281G, D320G, D457G) deleterious, deleterious, damaging, neutral/disease/ without meta-snp prediction/ disease; rs377266054→(E511V, E626V) deleterious, neutral, damaging, without meta-snp prediction /disease; rs377603311→(R844H, R883H, R721H, R707H, R506H) deleterious, deleterious /deleterious /neutral/deleterious /deleterious, damaging, disease/ disease/ disease/ without meta-snp prediction /disease; these represented sequential predictions for sift, sift & proven, sift in meta-snp ordinary.

PhD-SNP prediction for both SNPs & GO and Meta-snp tools showed similar prediction, in addition to some similarity or very tiny differences in PhD-SNP probability score.

SNPS & GO prediction, SNAP and Meta- SNP prediction were different in:

rs1145232→(G286R) Neutral, Disease, Disease; rs5742973→(E27Q) Neutral, Disease, Disease; rs56309301→(N243T) Neutral, Disease, Disease; rs111254723→(Y888C, Y927C) Neutral, Disease, Neutral; rs143010673→(W137C, W198C) Neutral, Neutral, Disease/ Neutral, Disease, Disease; rs143554211→(R259C) Neutral, Disease, Neutral; rs145103030→(R542C) Neutral, Disease, Disease; rs147901996→(N260S) Neutral, Disease, Neutral; rs185306467→(I136N, I197N) Disease, Neutral, Disease/ Neutral, Neutral, Disease; rs188947672→(L706V, L155V ) Neutral, Disease, Neutral; rs193252599→(E168K) Neutral, Disease, Neutral; rs200172149→(T202A) Neutral, Disease, Neutral; rs200244068→(D167G) Neutral, Neutral, Disease; rs200395679→(A104T ) Disease, Neutral, Neutral; →(L4S ) Neutral, Disease, Disease; rs368282144→(P514A, P475A) Neutral, Disease, Disease; rs370084230→(R10Q) Neutral, Disease, Neutral; rs370578519→(P661L, P700L, P639L) Neutral, Disease, Disease/ Neutral, Disease, Disease/ Neutral, Disease, Neutral; rs372752293→(T49R) Disease, Neutral, Disease; rs374222815→(V505M) Disease, Neutral, Neutral; rs374222815→(V843M,V882M,V720M) Disease, Neutral, Neutral; rs374879205→(P850A, P889A, P727A) Neutral, Disease, Disease; rs375553851→(N703H, N725H, N764H, N152H) Neutral, Disease, Disease/ Neutral, Disease, Disease/ Neutral, Disease, Disease/Neutral, Disease, Neutral; rs375640863→(D435G, D281G, D457G) Neutral, Disease, Disease; rs377266054→(E626V) Neutral, Disease, Neutral; rs377603311→(R506H) Disease, Disease, Neutral; those were not mention they are similar among the three prediction tools SNPS & GO prediction, SNAP and Meta- SNP prediction sequential.

In rs371999153→W49C was different from others SNPs because it’s responsible for ORMDL1 gene or (ORF Names: HSPC202) that located in the same chromosome position 2 (2:189782449) (https://www.ncbi.nlm.nih.gov/snp/?term=rs371999153) similar to PMS1 gene, their protein called ORM1-like protein 1 or Adoplin-1; responsible for negative regulator of sphingolipid synthesis (http://www.uniprot.org/uniprot/Q9P0S3),(https://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?rs=371999153), for Biological process is responsible for: (1) cellular sphingolipid homeostasis Source (2) ceramide metabolic process Source (3) negative regulation of ceramide biosynthetic process; they are widely expressed in adult and fetal heart, brain, lung, liver, skeletal muscle and kidney with expression in adult pancreas and placenta and in fetal spleen abdomen thymus while they expressed at intermediate level in pancreas, placenta and brain but low in skeletal muscle and lung, in addition they were found in subcellular location such as ERM (endoplasmic reticulum membrane) and multi-pass membrane protein.

The output of this study was explained and confirmed the damaging effect of those selected hemophilia B Snps.

References

1. Konkle AB, Josephson CN, Fletcher NS (2014) Hemophilia B. Initial Posting: Last Update.
2. Haemophilia B. From Wikipedia, the free encyclopedia.
3. Kniffin LC (2008) Hemophilia B. Online Mendelian Inheritance in Man, Johns Hopkins University.
4. Choi Y, Sims GE, Murphy S, Miller JR, Chan AP (2012) Predicting the functional effect of amino acid substitutions and indels. PLoS One 7: e46688.
5. Choi Y (2012) A fast computation of pairwise sequence alignment scores between a protein and a set of single-locus variants of another protein. In Proceedings of the ACM Conference on Bioinformatics, Computational Biology and Biomedicine, ACM, New York, USA, 414-417.
6. Choi Y, Chan AP (2015) PROVEAN web server: a tool to predict the functional effect of amino acid substitutions and indels. Bioinformatics 31: 2745-2747.
7. Adzhubei IA, Schmidt S, Peshkin L, Ramensky VE, Gerasimova A, et al. (2010) A method and server for predicting damaging missense mutations. Nat Methods 7: 248-249.
8. Adzhubei I, Jordan DM, Sunyaev SR (2013) Predicting functional effect of human missense mutations using PolyPhen-2. Curr Protoc Hum Genet 7: Unit7.20.
9. Ramensky V, Bork P, Sunyaev S (2002) Human non-synonymous SNPs: server and survey. Nucleic Acids Res 30: 3894-3900.
10. Sunyaev SR, Eisenhaber F, Rodchenkov IV, Eisenhaber B, Tumanyan VG, et al. (1999) PSIC: profile extraction from sequence alignments with position-specific counts of independent observations. Protein Eng 12: 387-394.
11. Capriotti E, Fariselli P, Casadio R (2005) I-Mutant2.0: predicting stability changes upon mutation from the protein sequence or structure. Nucleic Acids Res 33: W306-W310.
12. Capriotti E, Fariselli P, Calabrese R, Casadio R (2005) Predicting protein stability changes from sequences using support vector machines. Bioinformatics 21: ii54-ii58.
13. Capriotti E, Calabrese R, Casadio R (2006) Predicting the insurgence of human genetic diseases associated to single point protein mutations with Support Vector Machines and evolutionary information. Bioinformatics 22: 2729-2734.
14. Calabrese R, Capriotti E, Fariselli P, Martelli PL, Casadio R (2009) Functional annotations improve the predictive score of human disease-related mutations in proteins. Hum Mutat 30: 1237-1244.
15. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, et al. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389-3402.
16. Thomas PD, Kejariwal A (2004) Coding single-nucleotide polymorphisms associated with complex vs. Mendelian disease: Evolutionary evidence for differences in molecular effects. Proc Natl Acad Sci U S A 101: 15398-15403.
17. Capriotti E, Altman RB (2011) Improving the prediction of disease-related variants using protein three-dimensional structure. BMC Bioinformatics 12: S3.
18. Kabsch W, Sander C (1983) Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers 22: 2577-2637.
19. Thusberg J, Olatubosun A, Vihinen M (2011) Performance of mutation pathogenicity prediction methods on missense variants. Hum Mutat 32: 358-368.
20. Capriotti E, Altman RB, Bromberg Y (2013) Collective judgment predicts disease-associated single nucleotide variants. BMC Genomics 3: S2.
21. Pei J, Grishin NV (2001) AL2CO: calculation of positional conservation in a protein sequence alignment. Bioinformatics 17: 700-712.
22. Hall M, Frank E, Holmes G, Pfahringer B, Reutemann P, et al. (2009) The WEKA data mining software: an update. ACM SIGKDD Explorations 11: 10-18.