F1000ResearchF1000Research2046-1402F1000 Research LimitedLondon, UK10.12688/f1000research.53214.1Research ArticleArticlesDiscrimination of the SARS–CoV-2 strains using of coloured s-LASCA-imaging of GB-speckles, developed for the gene “S” nucleotide sequences
[version 1; peer review: 1 approved with reservations]
UlianovaOnegaFormal AnalysisWriting – Original Draft Preparation1SaltykovYuryData CurationFormal AnalysisInvestigationProject Administration2UlyanovSergeyConceptualizationFunding AcquisitionInvestigationMethodologyResourcesSoftwareSupervisionWriting – Original Draft PreparationWriting – Review & Editinga12ZaytsevSergeyData CurationInvestigationValidation2UlyanovAlexanderData CurationSoftwareWriting – Original Draft Preparation1FeodorovaValentinaConceptualizationFunding AcquisitionMethodologyProject AdministrationResourcesSupervisionWriting – Original Draft PreparationWriting – Review & Editinghttps://orcid.org/0000-0002-3827-407Xb2
Department of Medical Physics, Saratov State University, Saratov, Saratov, 410012, Russian Federation
Laboratory for Molecular Biology and NanoBioTechnology, Federal Research Center for Virology and Microbiology, Branch in Saratov, Saratov, Saratov, 410076, Russian Federationprof.sergey.ulyanov@outlook.comfeodorovav@mail.ru
A recent bioinformatics technique involves changing nucleotide sequences into 2D speckles. This technique produces speckles called GB-speckles (Gene Based speckles). All classical strategies of speckle-optics, namely speckle-interferometry, subtraction of speckle-images as well as speckle-correlometry have been inferred for processing of GB-speckles. This indicates the considerable improvement in the present tools of bioinformatics.
Methods
Colour s-LASCA imaging of virtual laser GB-speckles, a new method of high discrimination and typing of pathogenic viruses, has been developed. This method has been adapted to the detecting of natural mutations in nucleotide sequences, related to the spike glycoprotein (coding the gene «S») of SARS–CoV-2 gene as the molecular target.
Results
The rate of the colouring images of virtual laser GB-speckles generated by s-LASCA can be described by the specific value of R. If the nucleotide sequences compared utilizing this approach the relevant images are completely identical, then the three components of the resulting colour image will be identical, and therefore the value of R will be equal to zero. However, if there are at least minimal differences in the matched nucleotide sequences, then the value of R will be positive.
Conclusion
The high effectiveness of an application of the colour images of GB-speckles that were generated by s-LASCA- has been demonstrated for discrimination between different variants of the SARS–CoV-2 spike glycoprotein gene.
LASCASARS–CoV-2GB-specklesgeneRussian Foundation for Basic Research20-04-60349The reported study was funded by RFBR, project number 20-04-60349. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.Introduction
As it is well known, if laser light diffracts on random objects, then laser speckles are formed.
1-
3 Recently, the possibility of transforming a nucleotide sequence into a pattern of 2D speckles had been demonstrated.
4-
9 This new type of speckle pattern has been called “GB-speckles” (gene-based speckles).
5,
7,
9 Changes within in the structure of the GB-speckles can reflect even negligible changes in the nucleotide sequence, caused by inartificial mutations. This allows detection of single-nucleotide polymorphisms (SNP) using virtual GB-speckles with outstanding precision. In addition, it offers unlimited potential of improving the diagnosis’ accuracy by increasing the Fourier transform area.
10
Essential advancement in the area of GB-speckles has been reported in previous years. According to previously published reports,
4-
9,
11 implementation of speckle-optics methods, like speckle-interferometry and subtraction of speckle-images as well as speckle-correlometry for processing of GB-speckles, provides considerable progress in the current bioinformatics toolbox. This can become crucial to significantly improve existing routine methods of laboratory diagnostics of infectious diseases. GB-speckles as a technique opens the door to the new horizons in digital biology.
12,
13
Recently, model GB-speckle patterns of nucleotide sequences of the
omp1 genes for two different of
Chlamydia spp., such as
Chlamydia trachomatis and
Chlamydia psittaci of at least six genovars (D, E, F, G, J and K) have been composed.
4,
5 Probability density functions and correlation properties of spatial intensity fluctuations for the relevant GB-speckle patterns have been studied.
5-
7 As it has been shown in previous studies,
4-
7,
9 the presence of inartificial mutations in analysed strains, including single SNP cases, can be easily defined using methods of speckle-optics.
4-
7,
9 More recently, the encoding algorithm’s optimization for nucleotide sequences of
C. trachomatis into two-dimensional GB-speckle pattern had been carried out;
4,
6 and speckle-interferometric technique may give rise the ultra-fast optical processors of DNA sequences.
4 This is ensured by the development of the exclusive system of interferential fringes which are generated by the model interference pattern led by the existence of any type of mutations. Additionally, the method of virtual phase-shifting speckle-interferometry was reported to be efficacious
11 to investigate of polymorphism of the
C. trachomatis omp1 gene. This approach allowed the detection of the
C. trachomatis omp1 gene with SNPs, including both a single SNP and a combination of several SNPs in the bacterial strains with genetic mutations (11 known subtypes in total) had been developed.
6
The format of GB-speckles had been successfully applied to transform the nucleotide sequences of the genes expressing the serine proteases, the well-known Omptin family proteins within the
Enterobacteriaceae. These proteins have been found on the surface of several bacterial agents causing different enteric infections, such as salmonellosis, shigelosis, yersiniosis, and escherichiosis.
7 Further, the phase and the relevant two-dimensional distributions of the intensity of GB-speckles in various strains of viral pathogens, namely of lumpy skin disease virus of cattle, LSDV, and also for sheep-pox virus, SPPV have been obtained.
8 Additionally, interference patterns for generated the specific superposition in the relevant fields of GB-speckle and the certain difference in their images have been successfully investigated to reveal a minimal discrimination between the initial viral nucleotide sequences.
A new bioinformatics approach has been proposed very recently:
14 GB-speckles processing
via an
s-LASCA technique (from the
spatial
Laser
Speckle
Contrast
Analysis) application. As it had been demonstrated, it is possible to extend affectability of the proposed approach comparing to current bioinformatics strategies
15 using
s-LASCA imaging in the GB-speckles’ processing. It had been shown in Ref.
16, that the GB-speckles’ generation combined with
s-LASCA imaging method are very effective to analyze nucleotide polymorphism in several genes of
C. trachomatis.
This paper is devoted to development of advantageously new technique: the coloured
s-LASCA imaging of GB-speckles. Such a technique is an improved version of previously suggested “greyscale”
s-LASCA imaging that was recently developed especially for GB-speckles. Nucleotide sequences for some target genes SARS–CoV-2 have been successfully processed using coloured
s-LASCA-imaging. Natural mutations in the comparing genes have been reliably and accurately detected.
MethodsNucleotide sequences under consideration.
Seven nucleotide sequences of spike glycoprotein of SARS-CoV-2, namely:
the gene#1. hCoV-19/cat/USA/TX-TAMU-078/2020 (Accession ID: EPI ISL 699509),
the gene#2. hCoV-19/cat/Russia/RII-LEN-22246S/2021 (Accession ID: EPI ISL 811147),
the gene#3. hCoV-19/cat/Greece/2K/2020 (Accession ID: EPI ISL 717979),
the gene#4. hCoV-19/Wuhan/WIV04/2019 (Accession ID: EPI ISL 402124),
the gene#5. hCoV-19/England/QEUH-B11766/2020 (Accession ID: EPI ISL 642476),
the gene#6. hCoV19/South Africa/KRISP-EC-K005299/2020 (Accession ID: EPI ISL 678597),
the gene#7. hCoV-19/Russia/MOS-CRIE-13604226/2020 (Accession ID: EPI ISL 754198).
have been compared on the base of analysis of GB-speckles. The official reference sequences were taken from the
GISAID database.
Algorithm for the total conversion of a nucleotide sequence to a colour GB speckle structure, processed by
s-LASCA imaging technique
Initial processing of nucleotide sequence
First, the sequence of the letters derived from the original one-dimensional nucleotide sequence was converted into the sequence of numbers in accordance with the following rule:
4A→1;C→2;G→3;T→4.
It is critical to emphasize that the specific relationship between the letters and numbers in this case is not critical as used earlier;
6 thus, other rules could have been applied to the encoding, for instance:
C→1;G→2;T→3;A→4.or
T→1;A→2;C→3;G→4.
Next, all possible triad combination are generated. As a result, a complete set of all triads is formed:
(111),(112),(113),(114),(121),(122),(123),(124),(131),...,(444).
The number of all possible combinations of four numbers combined in triads is 64.
Then, a discrete magnitude, h, is allotted to each triad in accordance with the simple algorithm described previously.
4 This algorithm was implemented in Matlab R2015a (
RRID:SCR_001622); an open access alternative is
Julia. The value of h is a positive integer, varying in the range from 1 to 64. In this case, each triad from the original nucleotide sequence is associated with only one h value. So, for example, the combination (1 1 1) conforms to the value h = 1, (1 1 2) corresponds to h = 2, (1 1 3) conforms to h = 3, (1 1 4) conforms to h = 4, (1 2 1) conforms to h = 5, (1 2 2) conforms to h = 6, and so on. Finally, the latest combination (4 4 4) conforms to the value h = 64. Finally, a square matrix H
n,m was formed by a one-dimensional array h. The physical significance of the shaped matrix H
n,m is that each of its elements represents the local height of some virtual rough surface corresponding to the local content of the analyzed genetic construction. The resulting virtual rough surfaces could be used to model original speckle structures corresponding to diverse particular nucleotide sequences.
The two-dimensional speckle patterns that corresponded to each specific sequence was generated with the use the diffraction of a coherent beam with a square cross-section profile on a virtual scattering surface with a microrelief described by the matrix H
n,m. At each point of the virtual diffuser (in the beam scattering plane), some phase modulation U
n,m = exp(−2π
j H
n,m/64) is introduced (
j is an imaginary unit). The surface is illuminated at the normal incidence of the beam; the phase in the illuminating beam was a constant value.
Generating GB-speckles
The procedure for transcoding the original nucleotide sequence into a GB-speckle structure using the example of the hCoV-19/cat/USA/TX-TAMU-078/2020|2020-07-29 gene (the gene #1) is shown below.
After converting a sequence of letters into a sequence of numbers in accordance with the algorithm described by rule (1) described previously, the nucleotide sequence takes the following form:
As a result of diffraction of coherent beam on the phase screen (the virtual heights of the irregularities presented in the table (6)) with a square cross-section is formed GB-speckle-structure of two-dimensional intensity distribution, see
Figure 1a. Two-dimensional phase distribution GB, the speckle structure is shown in
Figure 1b.
GB-speckles: (a) Speckle pattern for 2D intensity distribution, b) Speckle pattern for 2D phase distribution.
<italic toggle="yes">s-LASCA</italic> imaging of GB-speckles
s-LASCA strategy has been connected for handling of GB-speckles. The strategy of
s-LASCA is based on the examination of an individual realization of static speckles.
3 In this case, the whole realization of the speckle field is divided into square zones; typically, each counting 5×5 or 7×7 pixels.
For each zone, the contrast of GB-speckles was calculated using the simplest formula:
C=σI〈I〉
where
I was the varying intensity of GB-speckles, changing from point to point;
σ
I was the standard deviation of the intensity of fluctuations. After the contrast
C is calculated in each point,
LASCA image is developed. Here, the size of subarea for the local contrast calculating was 2×2 pixels. As it has been demonstrated
14 this size of subarea is close to optimal.
Coloured GB speckles
To generate three two-dimensional implementations of GB speckles built for different genetic sequences, it is necessary to construct a colour image, where each colour component (red, green, and blue) has its own GB speckle structure. When all three speckle structures were totally indistinguishable, the colour images look grey-scale. If the colour components differ from each other, then, as a result, colouring will appear in the image.
In
Figure 2a, the coloured speckle-pattern for intensity distribution is presented (the red component obtained for the nucleotide sequence derived from gene #1, the green component corresponded to the nucleotide sequence of gene #2, and blue component was the relevant to gene #3 nucleotide sequence, respectively).
Coloured GB speckles: a) Coloured speckle-pattern for 2D intensity distribution, b) Coloured speckle-pattern for 2D phase distribution.
Figure 2a, demonstrates the differences in the initial nucleotide sequences, a slight staining appears in the colour speckle-pattern structure for a two-dimensional intensity distribution.
In
Figure 2b, the coloured speckle-pattern for phase distribution is shown for such nucleotide sequences as: (i) the red component for gene #1, greenfor gene #2, blue for gene #3.
It is quite obvious that in the case under consideration, there is a pronounced colouring over the whole image for the field of GB-speckle.
Thus, the obtained colour image for the intensity and phase of GB speckles is a reliable diagnostic sign of the presence of polymorphism.
A novel detection technique based on the
<italic toggle="yes">s-LASCA</italic> images with coloured GB-speckles
Once an
s-LASCA image is obtained for each of the three components of the matched genetic sequence, the final colour image can be constructed. An example of such an image is shown in
Figure 3a.
Coloured
<italic toggle="yes">s-LASCA</italic> images of GB-speckles, generated for three SARS–CoV-2 genes (the gene#1, the gene#2 and the gene#3).
Coloured
<italic toggle="yes">s-LASCA</italic> images of GB-speckles, generated for another three SARS–CoV-2 genes (the gene#4, the gene#5 and the gene#6).
Results and discussion
It is obvious that the image shown in
Figure 3a in comparison with the image in
Figure 2a has a more pronounced colouring over the entire field of view, but is characterized by a higher contrast. From a quantitative point of view, the degree of colouring can be described by the value
R=1N×M∑i=1N×M13×Iri−Ioi2+Igi−Ioi2+Ibi−Ioi2Ioi2
where
Iri,
Igi and
Ibi are values of intensity for the red, green, and blue components in each pixel,
Ioi=Iri+Igi+Ibi3
is the average intensity value in each pixel,
i is the pixel number,
M and
N are the number of rows and columns of the analyzed image,
N ×
M is the total number of pixels in the image.
Coloured
<italic toggle="yes">s-LASCA</italic> images of GB-speckles, generated for more three SARS–CoV-2 genes (the gene#5, the gene#6 and the gene#7).
Obviously, if the nucleotide sequences compared using
s-LASCA imaging of GB-speckles are completely identical, then the three components of the resulting colour image will be identical, and therefore the value of
R will be equal to zero. However, if there are at least minimal differences in the compared nucleotide sequences, then the value of
R will take a positive value. Thus, the value of
R calculated for the
Figure 3a is 0.1 (gene#1, gene#2 and gene#3 are compared).
In
Figure 3b, comparison of new SARS–CoV-2 genes: hCoV-19/Wuhan/WIV04/2019|2019-12-30 (gene#4), hCoV-19/England/QEUH-B11766/2020|2020-11-02 (gene#5) and hCoV19/South Africa/KRISP-EC-K005299/2020|2020-11-19 (gene#6) is presented.
R calculated for
Figure 3b equals to 0.596.
Finally, three SARS–CoV-2 genes are reflected in
Figure 3c (hCoV-19/England/QEUH-B11766/2020|2020-11-02 (gene#5), hCoV19/South Africa/KRISP-EC-K005299/2020|2020-11-19 (gene#6) and hCoV-19/Russia/MOS-CRIE-13604226/2020|2020-11-09 (gene#7). Again,
R equals to 0.596 for this case.
It is important to note that the value of
R calculated for
Figure 2a and
Figure 2b (coloured bare GB-speckle) equals to 0.049 and 0.026, respectively. This means that the value of
R at least in two times higher for GB speckles, processed by
s-LASCA imaging technique.
Evidently,
R is positive for all images in
Figures 3; so,
R is an important diagnostic feature when detecting the presence of SNPs in SARS–CoV-2 genes. This is the main result.
Conclusion
A fundamentally new bioinformatics technique for reliable detection of single SNPs is proposed. The new method is based on the applying of the
s-LASCA ‘imaging technique’ generating original GB-speckles. It is established that even one SNP can be reliably detected. It has been demonstrated that suggested technique is very effective tool for discrimination between different variants of the SARS–CoV-2 spike glycoprotein gene.
Data availabilityUnderlying data
GISAID Gene: hCoV-19/cat/USA/TX-TAMU-078/2020. Accession number EPI ISL 699509;
GISAID Gene: hCoV-19/cat/Russia/RII-LEN-22246S/2021. Accession number EPI ISL 811147;
GISAID Gene: hCoV-19/cat/Greece/2K/2020. Accession number EPI ISL 717979;
GISAID Gene: hCoV-19/Wuhan/WIV04/2019. Accession number EPI ISL 402124;
GISAID Gene: hCoV-19/England/QEUH-B11766/2020. Accession number EPI ISL 642476;
GISAID Gene: hCoV19/South Africa/KRISP-EC-K005299/2020. Accession number EPI ISL 678597;
GISAID Gene: hCoV-19/Russia/MOS-CRIE-13604226/2020. Accession number EPI ISL 754198.
Sequences are available after registration at the
GISAID public database.
ReferencesDaintyC:
Laser speckle and related phenomena.Topics in Applied Physics.1984;9.
18213052AizuY:
Biospeckle method for retinal blood flow analysis: flexible correlation measurements.Optical Diagnostics of Biological Fluids and Advanced Techniques in Analytical Cytology.SPIE Press;1997;91–102.BriersJD:
Laser Doppler, speckle and related techniques for blood perfusion mapping and imaging.Physiol Meas.2001;22:R35–R66.
1176108110.1088/0967-3334/22/4/201UlyanovSS:
Using of methods of speckle optics for
Chlamydia trachomatis typing.Proc. SPIE.2017;10336:103360D.UlyanovSS:
Statistics on gene-based laser speckles with a small number of scatterers: implications for the detection of polymorphism in the
Chlamydia trachomatis omp1 gene.Laser Physics Letters.2018;15.
10.1088/1612-202X/aaa11cFeodorovaVA:
Optimization of algorithm of coding of genetic information of
Chlamydia.Proc. SPIE.2018;10716:107160Q.
10.1117/12.2314640UlyanovSS:
Study of statistical characteristics of GB-speckles, forming at scattering of light on virtual structures of nucleotide gene sequences of Enterobacteria.Izv. Saratov Univ. (N. S.), Ser. Physics.2018;18:123–137.
10.18500/1817-3020-2018-18-2-123-137SaltykovYV:
Analysis of nucleotide sequences of the gene GPCR of the genus Capripoxvirus representatives using speckle interferometry of GB speckles and subtracting their images.Math. mod., comp. and natural exp. in natural sci.2020:2.UlianovaOV:
Speckle-interferometry and speckle-correlometry of GB-speckles.Front Biosci (Landmark Ed).2019;24:700–711.
30844706GoodmanJW:
Introduction to fourier optics.
New York:
McGraw Hill Companies;1988.FeodorovaVA:
Application of virtual phase-shifting speckle-interferometry for detection of polymorphism in the
Chlamydia trachomatis omp1 gene.Proc. SPIE.2018;10716:107160M.
10.1117/12.2314700SleatorRD:
Digital biology. A new era has begun.Bioengineered.2012;3:311–312.
2309945310.4161/bioe.22367PMC3489704FreddolinoPLTavazoieS:
The Dawn of Virtual Cell Biology.Cell.2012;150:248–250.
2281788810.1016/j.cell.2012.07.001PMC4430847UlianovaOV:
LASCA-imaging of gene-based speckles: application to detection of the gene polymorphism in a bacterial model.Laser Physics Letters.2020;17.
10.1088/1612-202X/ab8b66LeskAM:
Introduction to BioinformaticsOxford University Press;2002.UlianovaOV:
Could
LASCA-imaging of GB-speckles be applied for a high discrimination and typing of pathogenic bacteria?PloS one.2021;16.10.5256/f1000research.56573.r88324Reviewer response for version 1AngelskyOleg1Refereehttps://orcid.org/0000-0002-5463-8961
Department of Correlation Optics, Chernivtsi National University, Chernivtsi, Ukraine
Competing interests: No competing interests were disclosed.
The paper is devoted to the investigations of virtual Gene Based speckles for the determination the difference between SARS-Cov-2 spike glycoprotein genes. As it was noted by the authors the given method can use the classical approaches of speckle-optics. But there are some questions:
Any optical speckle is the result of the interference of waves scattered by inhomogeneities, and to obtain a speckle pattern, the wavelength is decisive. Thus, the question arises about the concrete wavelengths used in the study and their relationships with pseudo inhomogeneities. It would be interesting to know about the coherence of the sources used the modelling.
Please present an experiment setup, and the results of experimental modelling.
It would be interesting to conduct a complete analysis of the entered parameter R, to represent its physical, biochemical meaning. The virus is constantly mutating, how is it possible to estimate the degree of virus mutation using this parameter.
The paper can be indexed after reworking.
Is the work clearly and accurately presented and does it cite the current literature?
Partly
If applicable, is the statistical analysis and its interpretation appropriate?
Yes
Are all the source data underlying the results available to ensure full reproducibility?
Yes
Is the study design appropriate and is the work technically sound?
Yes
Are the conclusions drawn adequately supported by the results?
Partly
Are sufficient details of methods and analysis provided to allow replication by others?
Yes
Reviewer Expertise:
Singular and Correlation Optics
I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
FeodorovaValentinaFederal Research Center for Virology and Microbiology, Branch in Saratov, Russian Federation
Competing interests: No competing interests were disclosed.
3072021
Reply to the comments of reviewer Prof. Oleg Angelsky:
Any optical speckle is the result of the interference of waves scattered by inhomogeneities, and to obtain a speckle pattern, the wavelength is decisive. Thus, the question arises about the concrete wavelengths used in the study and their relationships with pseudo inhomogeneities. It would be interesting to know about the coherence of the sources used the modelling.
Response: This is really reasonable and very important question. Fragment below is added to the new version of the text:
It is assumed that speckles are formed in the far diffraction zone and described in the Fraunhofer approximation. In this case, the expression for the amplitude of the scattered field is the Fourier transform of the field in the diffraction plane, evaluated at frequencies spaces
Fx=Xo/( z* λ), Fy=Yo/( z* λ), (*)
where Xo and Yo are the coordinates in the observation plane, z is the distance between the scattering plane and the observation plane, λ is the wavelength. The illuminating radiation is completely monochromatic, thus, λ =const.
Reference to Goodman, J. W. Introduction to fourier optics.
McGraw Hill Companies, New York (1988).
In this situation, the structure of speckles does not depend on the wavelength and z. Only the sizes of GB-speckles depend on these values, the average size of which is determined by the ratio:
d~3*z* λ /a (**)
where a is the size of the illuminated fragment of virtual surface. It is important to emphasize that the ratio λ/a characterizes the diffraction angular divergence of a laser beam in the far field, and the product of this divergence angle by the light traveled distance z is equal to the lateral size of the beam.
New reference: M. Francon. La granularite laser (speckle) et ses applications en optique. Masson, Paris, New York, Barcelone, Milan 1978.
Thus, it can be seen that the diameter of the undisturbed laser beam (namely, this value is on the right side of the expression (**)) and the average speckle size are approximately equal to each other in any observation plane.
In other words, when the parameters z and λ change, a proportional change in the size of all speckles occurs synchronously. At the same time, the structure of speckle-patterns in all observation planes are completely similar, only their scale changes from plane to plane, but not the shape of the speckles or their location in the speckle pattern.
Please present an experiment setup, and the results of experimental modelling.
Response: Experimental studies were not carried out in this work, only computer modeling. The scheme for calculating GB speckles during radiation diffraction on a virtual scattering surface is described in detail in the work: Ulianova OV, et al.: Speckle-interferometry and speckle-correlometry of GB-speckles. Frontiers in Bioscience-Landmark 2019; 24, 700-711. This fact is mentioned in the new version of this article.
It would be interesting to conduct a complete analysis of the entered parameter R, to represent its physical, biochemical meaning. The virus is constantly mutating, how is it possible to estimate the degree of virus mutation using this parameter.
Response: The physical meaning of the introduced parameter R is that this parameter characterizes the degree of coloring of the picture (GB-speckle- pattern). The bioinformatic (molecular biology) value of R is that it takes positive values, even in the case of the appearance of a one SNP in the analyzed nucleotide sequences. Thus, the minimum natural mutations of the virus can be determined using the parameter R. This circumstance is also noted in the Conclusions of new version of this article.
The authors are extremely grateful to the respected reviewer, whose critical comments have significantly improved the quality of this article.