Quantification of the Diversity in Gene Structures Using the Principles of Polarization Mapping

Round 1

Reviewer 1 Report

In this paper, Dmitry et. al propose and develop a two-dimensional screen based approach for displaying and analyzing the diversity of gene sequence “structures”.  In general, the manuscript is well-presented, although the English language and style needs to be improved. The idea of the paper is interesting. However, I am concerned the results are not very convincing/exciting for the bioinformatics community. Thus, the significance and impact of the paper is unclear/limited. I would suggest the authors rewrite the paper. Additional experiments, datasets and case studies could be helpful.

1,

The English writing of the paper is awkward, and sometimes cumbersome. It also has some grammar issues. Like,

[Line 15]“A possibility for displaying and analyzing a diversity of gene structures using polarization coding is considered”

Extensive editing and proofreading are needed.

2,

Source code should be made public, like posting it on GitHub. Otherwise, it will be difficult for the bioinformatics community to adopt the idea.

3,

Some more technical details are needed. For example,

[line 211-212] “The scale factor Ksc was set equal to 0.1 in order to refine the features of distributions in the paraxial zone. Note that the used phase modulation algorithm”

It is not clear how a scale factor of 0.1 could “refine the features of distribution”. The authors should provide some guidelines or rule of thumb for parameter setting.

4,

[line 242-243] “diffraction pattern allows us to suggest that positions of the extreme values of s3k,m which are close to 1 or -1, are very sensitive to the changes in the structure.”

The authors mostly talk about the sequence composition changes or nucleotide changes in the sequence. It has nothing to gene’s 3D structure and the resulted protein structure. “Structure” would be confusing or misleading for bioinformatics researchers. Such sentences need to be rephrased.

5,

I think the idea of instrumental implementation of polarization encoding of the gene-based (GB) synthesized matrices is far away from reality. I don’t think this can be easily adapted into a typical DNA/RNA sequencing experiment. (Or the authors only separate experiment setting?) Is such implementation applicable? (cost, feasibility wise?)  Such idea should be included in the discussion or future work.

6,

Some paragraphs in the Discussion could be shortened, like the first paragraph. I suggest the authors add some discussion about the difference between the proposed approach and conventional approaches. How it will complement current Bioinformatics toolsets. Under what circumstance, the proposed approach is suitable for analysis. Otherwise, the significance of this paper will be quite limited.

Author Response

The authors are grateful to the reviewer for valuable comments that help improve the quality of the work. The reviewer's comments and suggestions are given below along with the answers of the authors.

1. “The English writing of the paper is awkward, and sometimes cumbersome. It also has some grammar issues. Like, [Line 15]“A possibility for displaying and analyzing a diversity of gene structures using polarization coding is considered”

Extensive editing and proofreading are needed.”

English writing has been edited. Long sentences were shortened or divided into pairs of separate sentences. The revised text was carefully proofread and edited by our English-experienced colleague.

2. “Source code should be made public, like posting it on GitHub. Otherwise, it will be difficult for the bioinformatics community to adopt the idea.”

The source codes with necessary comments have been presented in the Supplementary Materials to this work.

3. “Some more technical details are needed. For example, [line 211-212] “The scale factor Ksc was set equal to 0.1 in order to refine the features of distributions in the paraxial zone. Note that the used phase modulation algorithm”

It is not clear how a scale factor of 0.1 could “refine the features of distribution”. The authors should provide some guidelines or rule of thumb for parameter setting.”

Additional text and supporting graphics are included in the revised version (see pages 6 and 7, beginning with the words "Note that …." and ending with the words "…. entire analyzed area with a decrease in K_sc"; Figure 2).

4. “[line 242-243] “diffraction pattern allows us to suggest that positions of the extreme values of s3k,m which are close to 1 or -1, are very sensitive to the changes in the structure.”

The authors mostly talk about the sequence composition changes or nucleotide changes in the sequence. It has nothing to gene’s 3D structure and the resulted protein structure. “Structure” would be confusing or misleading for bioinformatics researchers. Such sentences need to be rephrased”.

It has been corrected.

5. “I think the idea of instrumental implementation of polarization encoding of the gene-based (GB) synthesized matrices is far away from reality. I don’t think this can be easily adapted into a typical DNA/RNA sequencing experiment. (Or the authors only separate experiment setting?) Is such implementation applicable? (cost, feasibility wise?) Such idea should be included in the discussion or future work.”.

The authors do not agree with the reviewer that the considered method of analysis and visualization of genetic information is “far away from reality.” It quite admits instrumental implementation; this point is discussed in the "Discussion" section (see page 19, beginning with the words "As an example of possible instrumental implementation ….." and ending with the words "…. can be used in combination with an appropriate objective lens."). Moreover, the instrumental implementation of the considered method is the subject of our further work. In addition, this approach is not intended to be integrated into DNA-sequencing systems. It is proposed for use at the second stage of DNA analysis (analysis and visualization of DNA-associated symbol sequences, see “Introduction” section).

6. “Some paragraphs in the Discussion could be shortened, like the first paragraph. I suggest the authors add some discussion about the difference between the proposed approach and conventional approaches. How it will complement current Bioinformatics toolsets. Under what circumstance, the proposed approach is suitable for analysis. Otherwise, the significance of this paper will be quite limited.”

Discussion has been shortened; the first paragraph has been removed. We believe, with respect to another remark, that the place and possibilities of the discussed hybrid technique among computational approaches in bioinformatics have been discussed in the “Introduction” (see page 3, beginning with the words "At the same time, the problem ….." and ending with the words "…. computer solutions to bioinformatics problems."

7. “Additional experiments, datasets and case studies could be helpful.”

An additional text fragment has been added to the "Discussion" section (see pages 18, 19 of the revised manuscript, beginning with the words "Note that a pilot computer simulation …" and ending with the words "… in the number of substitutions."). This fragment presents a discussion of the earlier pilot application of the considered polarization technique to various strains of the African swine fever (ASF) virus. A corresponding link to our recent publication regarding the pilot polarization analysis of symbol sequences for ASF strains is given in the updated list of references ([37], open access, doi: 10.18287/JBPE22.08.040302).

Author Response File: Author Response.docx

Reviewer 2 Report

The authors demonstrated nucleotide screening by their proposed polarization coding method. The local configuration of gene can be detected by the local polarization state (in particularly, S₃ in Stokes vector). Such a technique might potentially be applicable for gene screening in the future. Thus, I believe the manuscript deserve publication in Current issues in molecular biology. However, for reader’s understanding, I suggest that the author should add and improve some contents in their manuscript before the publication. For more detail, please see each comment.

(Comment #1) Information on the used laser is lacking. The authors should add information on laser such as continuous laser or pulsed laser, wavelength of laser, optical power illuminated onto gene samples, and so on.

(Comment #2) What is the typical value of the spatial resolution in local polarization measurements? And, the achieved spatial resolution is enough good to accurately screen gene configuration? The author should add the discussion on these points.

(Comment #3) How to interpret the binary maps in Figs. 3, 4(b), and 6(a, b)? The different gene sequences in the three (Wuhan, Delta, and Omicron) are summarized in table 1. So, I guess, the acquired binary maps include the information on the different gene sequences. But, I could not understand it. The author should add the explanation on the relation between binary maps and difference in gene sequences summarized in the table.

Author Response

The authors are grateful to the reviewer for valuable comments that help improve the quality of the work. The reviewer's comments and suggestions are given below along with the answers of the authors.

1. “Information on the used laser is lacking. The authors should add information on laser such as continuous laser or pulsed laser, wavelength of laser, optical power illuminated onto gene samples, and so on.”
2. “What is the typical value of the spatial resolution in local polarization measurements? And, the achieved spatial resolution is enough good to accurately screen gene configuration? The author should add the discussion on these points.”

The revised version includes an additional paragraph discussing the possible design of a polarimeter for the analysis of gene sequences (see page 19, beginning with the words "As an example of possible instrumental implementation ….." and ending with the words "…. can be used in combination with an appropriate objective lens."). In this paragraph, the main characteristics of basic optical elements (such as laser, spatial light modulator, CMOS camera) are discussed. In addition, a discussion of the relationship between the scale factor and the parameters of the system elements in instrumental implementation is given in Section 2 (see pages 6, 7, beginning with the words “Note that ..” and ending with the words “… in the case of instrumental implementation)”.   

3. “How to interpret the binary maps in Figs. 3, 4(b), and 6(a, b)? The different gene sequences in the three (Wuhan, Delta, and Omicron) are summarized in table 1. So, I guess, the acquired binary maps include the information on the different gene sequences. But, I could not understand it. The author should add the explanation on the relation between binary maps and difference in gene sequences summarized in the table.”

A binary map of extreme local polarization states represents a unique two-dimensional identifier for a given sequence. The positions of the bright dots on the map are uniquely determined by the distribution of the four base nucleotides (A,C,T,G) over the sequence and change when a portion of the nucleotides is replaced. It is possible to draw analogies with a hologram, which is not a direct image of an object in the generally accepted sense, but at the same time contains complete information about its structure and shape. As shown in our work, changes in binary maps due to changes in the structures of nucleotide sequences can be quantified using the coefficient of their mutual correlation. An additional comment on this point has been added to the revised version (see page 11, the last sentence). In addition, some comments on this subject were already present in the text (see, e.g., pages 9 and 10, the first paragraph in subsection 2.3).

Author Response File: Author Response.docx

Reviewer 3 Report

The authors presented novel computational method of DNA sequence visualization on 3D plot called polarization mapping. The application of the sequence variability visualization is shown on the examples of coronavirus sequences. The work is novel, interesting, and the results are supported by the computational study.

However, presentation style is somehow misleading. The references are given in bulk. Some important work in this area should be cited in the text.

Remarks on the text.

Some clear wording about computation approach should be given in the Abstract and might be in the paper title (“computational polarization methos” , “computational method”, “visualization of DNA sequence” or like that)

Line 15: Change word “possibility” to computer method”.

Line 16: change  “considered” to “presented”

Line 22: ‘etc.’ – remove, or list all the analysis variants.

Line 25: “various approaches to…” – change to “The application of…”

Line 26: Words about “instrumental implementation” is unclear. Give details, or change this phrase.

Line 29: - keywords should be updated. Add some general terms such as “bioinformatics”, “sequence analysis”, and some words about applications such as “coronavirus genome”, “genome structure visualization”

Major remark is on the in-text citation style – bulk references such as “etc. [1-10].” and “[11-14]” are not appropriate. Need cite papers separately, one-two, may three references together. Need rephrase, add details, give references after each sentence.

What references from 1-10 are for “correlation, and spectral analysis, neural network approaches, artificial…” methods?

Moreover, some common bioinformatics references such standard  BLAST are not relevant.

Rewrite text in lines 47-50.

I suggest adding proper references on sequence visualization in 2D and 3D space, sequence encoding methods.

Line 51: “the problem of gene diversity recognition and imaging” – change this phrase.

It is rather problem of gene difference analysis and visualization of such difference.

Line 73 and 76: “indices i,j” – need note the range, assume i=1,…N? Then what is N? just arbitrary integer number to describe range?

Line 86: “[18-21]” – several references together, cite separately. Reference number 18 cited before 15 (see line 95)

Line 95 – cited 15-17 – before 18. Please reorder. Avoid bulk citations.

Line 125: “N = n ** 2” - please indicate minimal values, assume N should be at least 16?

It is worthy make small figure instead of formula (1)

Some mathematical formalism is redundant. Function f( ) seems be function of single argument, but in formula (4) we see f(i,j).

Line 160-61: numbering “(1)” and “(5)” could be mixed with formula or reference. Please use “panel 1”, “panel 5 of Figure” or like that.

Figure 1 is interesting, but need write wording “computer” or “computational” to “The scheme…” It looks like experimental equipment, but it is in fact computer scheme of data processing.

Line 205: “As an example, consider..” – need add other examples, other viruses of genome sequences.. Only Wuhan stain has limited interest – there are millions of sequences known in the databases. Let state the problem more wide – to analyze differences and visualize any DNA sequence.

Line 230: “Spike gene” – please indicate gene ID, official gene nomenclature, proved a reference.

Line 249: Formula (7) – it sems some typo (signs “>(<)”) – two rows of the formula are identical.

Line 279: “(30 different triplets)” please indicate homology level, number of mismatches in standard alignment of these sequence.

Table 1 could be shorter. It is too technical. Need show main difference, or indicate ‘hot point’ of mutations

Line 448: “etc” – remove or add details about the applications. Currently the application fields are too common.

Line 524: “Not applicable.” – I think need indicate the database accession here.

In general it is interesting work, but too narrow.

Need extend the description of known works in the fields of DNA sequence coding, 2D plot construction, mention known approaches, such as complexity and entropy of DNA, DNA curves, chaos game and fractal presentations.

Here is list of suggest work to mention in background and discussion

(PubMed links and comments)

https://pubmed.ncbi.nlm.nih.gov/15215465/ (web-server on DNA complexity)

https://pubmed.ncbi.nlm.nih.gov/35181687/

(SciRep 2022 – more recent application)

https://pubmed.ncbi.nlm.nih.gov/15130826/

A.Bolshoy, 2003

https://pubmed.ncbi.nlm.nih.gov/12744704/

G. Gordon, 2003 Multi-dimensional linguistic complexity – again one of the beginning works

Musical application (side way – may not cite, just to know existing applications) (2019)

https://pubmed.ncbi.nlm.nih.gov/33267383/

https://pubmed.ncbi.nlm.nih.gov/24278218/

Armando J Pinho et al. 2013 (review)

V.D.Gusev et al. 1999 On the complexity measures of genetic sequences (first mathematical estimates of DNA coding)

Visualization of linear DNA sequence in 2D plots:

Maxim Pyatkov , Anton Pankratov . Bioinformatics (2014) SBARS: fast creation of dotplots for DNA sequences on different scales using GA-,GC-content

https://pubmed.ncbi.nlm.nih.gov/34465942/

(Review on previous 2D sequence visualization technique

Numerical characterization and similarity analysis of DNA sequences based on 2-D graphical representation of the characteristic sequences.

Li C, Wang J. Comb Chem High Throughput Screen. 2003. doi: 10.2174/138620703771826900.

C-curve: a novel 3D graphical representation of DNA sequence based on codons

Math Biosci. 2013

https://pubmed.ncbi.nlm.nih.gov/23246806/

 Recent review on 2D and 3D sequence presentation

Ashesh Nandy  (2022)

Comb Chem High Throughput Screen. 2022;25(3):354-364.

doi: 10.2174/1386207324666210510164743.

Mapping Biomolecular Sequences: Graphical Representations - Their Origins, Applications and Future Prospects

https://pubmed.ncbi.nlm.nih.gov/33970841/

https://pubmed.ncbi.nlm.nih.gov/20969878/

J Theor Biol. 2011 Jan 21;269(1):123-30.

doi: 10.1016/j.jtbi.2010.10.018.

Three 3D graphical representations of DNA primary sequences based on the classifications of DNA bases and their applications

Guosen Xie  & Zhongxi Mo

DNA chaos game conception

For review cite –

Comput Struct Biotechnol J . 2021

doi: 10.1016/j.csbj.2021.11.008.

Chaos game representation and its applications in bioinformatics

Recent similar woks -

Int J Mol Sci . 2022 doi: 10.3390/ijms23031847.

Fractal Analysis of DNA Sequences Using Frequency Chaos Game Representation and Small-Angle Scattering

Background works:

The existence of fractal sets of DNA sequences

Fractality of DNA texts.

Borovik AS, Grosberg AYu, Frank-Kamenetskii MD. J Biomol Struct Dyn. 1994. doi: 10.1080/07391102.1994.10508765.

Need extend discussion on coronavirus genome, detection of stain differences. Cite works in this field based on DNA encoding and various forms of presentation –

Coronavirus -

Similarity Studies of Corona Viruses through Chaos Game Representation.

Sengupta DC, Hill MD, Benton KR, Banerjee HN. Comput Mol Biosci. 2020 Sep;10(3):61-72. doi: 10.4236/cmb.2020.103004.

Recent work by 2022 - Detection of intra-family coronavirus genome sequences through graphical representation and artificial neural network

https://pubmed.ncbi.nlm.nih.gov/35095217/

https://pubmed.ncbi.nlm.nih.gov/34465942/

Nonlinear Dyn. 2021;106(2):1525-1555.

doi: 10.1007/s11071-021-06836-y. Epub 2021 Aug 27.

Advances in the computational analysis of SARS-COV2 genome

J A Tenreiro Machado (2021)

The list of references is not obligatory.

It is important rewrite introduction and cite previous work on DNA sequence presentation.

It will be actual add recent citation on coronavirus and extend application to other model sequences.

Author Response

The authors are grateful to the reviewer for valuable comments that help improve the quality of the work. The reviewer's comments and suggestions are given below along with the answers of the authors.

1. “Major remark is on the in-text citation style – bulk references such as “etc. [1-10].” and “[11-14]” are not appropriate. Need cite papers separately, one-two, may three references together. Need rephrase, add details, give references after each sentence. ….. What references from 1-10 are for “correlation, and spectral analysis, neural network approaches, artificial…” methods?

Moreover, some common bioinformatics references such standard  BLAST are not relevant.

Rewrite text in lines 47-50.

I suggest adding proper references on sequence visualization in 2D and 3D space, sequence encoding methods……

Need extend the description of known works in the fields of DNA sequence coding, 2D plot construction, mention known approaches, such as complexity and entropy of DNA, DNA curves, chaos game and fractal presentations.”

It has been done (see the updated “Introduction” section, pages 1-3, beginning with the words "Analysis of the genetic structure …" and ending with the words "…. belonging to representing polygonal lines", and the updated reference list, from [1] to [27]).

2. “It is important rewrite introduction and cite previous work on DNA sequence presentation. It will be actual add recent citation on coronavirus and extend application to other model sequences”.

It has been done. The "Introduction" section has been substantially rewritten (see our response to reviewer's comment #1). Additional references to recent research on the genetic structure of the coronavirus have been added to the updated list of references ([20], [21]). An additional text fragment has been added to the "Discussion" section (see pages 18, 19 of the revised manuscript, beginning with the words "Note that a pilot computer simulation …" and ending with the words "… in the number of substitutions."). This fragment presents a discussion of the earlier pilot application of the considered polarization technique to various strains of the African swine fever (ASF) virus. A corresponding link to our recent publication regarding the pilot polarization analysis of symbol sequences for ASF strains is given in the updated list of references ([37], open access, doi: 10.18287/JBPE22.08.040302).

The detailed comments

3. “Some clear wording about computation approach should be given in the Abstract and might be in the paper title (“computational polarization methos” , “computational method”, “visualization of DNA sequence” or like that)

Line 15: Change word “possibility” to computer method”.

The authors should note that the presented method is not purely a computational technique, but allows for future instrumental implementation in the form of a prototype of a polarimetric system (see more detailed comments in point 5).

4. “Line 16: change “considered” to “presented””.

It has been done.

5. Line 22: “‘etc.’ – remove, or list all the analysis variants”.

It has been done.

6. Line 25: “various approaches to…” – change to “The application of…”

It has been done.

7. Line 29: - keywords should be updated. Add some general terms such as “bioinformatics”, “sequence analysis”, and some words about applications such as “coronavirus genome”, “genome structure visualization”

It has been done.

8. Line 51: ““the problem of gene diversity recognition and imaging” – change this phrase.

It is rather problem of gene difference analysis and visualization of such difference.”

It has been done.

9. Line 73 and 76: ““indices i,j” – need note the range, assume i=1,…N? Then what is N? just arbitrary integer number to describe range?”

As follows from the technique for the synthesis of square phase-modulating matrices, the values of (i, j) vary from 1 to 2n. Here n is defined by the condition N = n*n. In turn, N defines the number of triplets in the analyzed fragment and is equal to the maximum square of an integer that is equal to or less than N_t/3 (N_t is the number of symbols in the analyzed sequence). 

9. Line 86: “[18-21]” – several references together, cite separately. Reference number 18 cited before 15 (see line 95)

Line 95 – cited 15-17 – before 18. Please reorder. Avoid bulk citations.

Bulk citations have been removed.

10. “Line 125: “N = n ** 2” - please indicate minimal values, assume N should be at least 16?

It is worthy make small figure instead of formula (1).

Some mathematical formalism is redundant. Function f( ) seems be function of single argument, but in formula (4) we see f(i,j)”.

The N value of 16 corresponds to 48 (A,C,T,G) symbols in the analyzed fragment of sequence. The extreme value of N is 1; in this case a single triplet is encoded and the synthesized matrix consists of only one 2*2 submatrix. This is trivial and is of no practical interest. In practice, the analyzed fragments of symbol sequences contain from several hundred to several thousand symbols obtained as a result of DNA sequencing. Accordingly, the number of rows and columns in the synthesized matrices typically ranges from two dozen to one hundred or more.

The authors note that there is no redundancy in writing Eqs. (3) and (4). In the general case, the principle of polarization coding under discussion allows various coding options, including a separate setting of the scaling phase factor and retardation depth for each element of the synthesized matrix. Accordingly, the value of the function f( ) describing retardation for some element of the matrix is determined by the position of this element in the matrix, that is, the set of values (i,j). A small figure illustrating the principle of coding is in the upper part of Fig. 1.

11. Line 230: “Spike gene” – please indicate gene ID, official gene nomenclature, proved a reference.

The processed sequences have been identified by the references in the reference list.

12. Line 249: Formula (7) – it sems some typo (signs “>(<)”) – two rows of the formula are identical.

This is not a typo. You should pay attention to the different mutual order of the symbols > and < in the first and second lines of formula 7 (>(<) against <(>)). The normalized components of the Stokes vector vary between -1 and 1. Accordingly, the symbols > (the first line) and < (the second line) correspond to the specified positive threshold value. In contrast, the opposite order (</>) corresponds to the specified negative threshold. The corresponding explanation is included in the text.

13. Line 279: “(30 different triplets)” please indicate homology level, number of mismatches in standard alignment of these sequence.

The additional information has been presented in the Supplementary Materials to this work (S3 section).

14. Table 1 could be shorter. It is too technical. Need show main difference, or indicate ‘hot point’ of mutations

The authors do not consider it necessary to reduce Table 1. The table has been edited by color highlighting matching (green) or different (red) triplets for strains “Delta” and “Omicron” with respect to strain “Wuhan” (as a reference).

15. Line 524: “Not applicable.” – I think need indicate the database accession here.

“Data Availability Statement” has been updated.

16. “Line 26: Words about “instrumental implementation” is unclear. Give details, or change this phrase…… Figure 1 is interesting, but need write wording “computer” or “computational” to “The scheme…” It looks like experimental equipment, but it is in fact computer scheme of data processing.”

The authors do not agree with the reviewer that the considered method of analysis and visualization of genetic information is “in fact computer scheme of data processing.” It quite admits instrumental implementation; this point is discussed in the "Introduction" section (see page 3, beginning with the words "At the same time, the problem …." and ending with the words "…. computer solutions to bioinformatics problems"). Additional discussion of the possibility of instrumental implementation is included in the "Discussion" section (see page 19, beginning with the words "As an example of possible instrumental implementation ….." and ending with the words "…. can be used in combination with an appropriate objective lens."). Moreover, the instrumental implementation of the considered method is the subject of our further work.

Author Response File: Author Response.docx

Reviewer 4 Report

Comments for "Quantification of the Diversity in Gene Structures Using the Principles of Polarization Mapping",

Authors did comprehensive work with multiple encoding methods, the test and evaluation are appropriate.

The discussion of the parameters is satisfactory and sufficient.

The paper is interesting to me as a scientist working in the artificial intelligence-related field.

Author Response

The authors are grateful to the reviewer for the high appreciation of their work.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

improved, can be accepted. 

Reviewer 3 Report

Thanks for the revision and detailed answer. I have no more critical remarks on the text.

Please only check actual access date for the references 34, 35, 36 - currently "(accessed on 15.08.2021)".