Citation: Planat, M.; Amaral, M.M.;
Irwin, K. Algebraic Morphology of
DNA–RNA Transcription and
Regulation. Symmetry 2023, 15, 770.
https://doi.org/10.3390/sym15030770
Academic Editor: Masato Yoshizawa
Received: 21 February 2023
Revised: 8 March 2023
Accepted: 15 March 2023
Published: 21 March 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
symmetry
S S
Article
Algebraic Morphology of DNA–RNA Transcription
and Regulation
Michel Planat 1,*,†
, Marcelo M. Amaral 2,†
and Klee Irwin 2,†
1 CNRS, Institut FEMTO-ST, Université de Franche-Comté, F-25044 Besançon, France
2 Quantum Gravity Research, Los Angeles, CA 90290, USA
* Correspondence: michel.planat@femto-st.fr
†
These authors contributed equally to this work.
Abstract: Transcription factors (TFs) and microRNAs (miRNAs) are co-actors in genome-scale de-
coding and regulatory networks, often targeting common genes. To discover the symmetries and
invariants of the transcription and regulation at the scale of the genome, in this paper, we introduce
tools of infinite group theory and of algebraic geometry to describe both TFs and miRNAs. In TFs,
the generator of the group is a DNA-binding domain while, in miRNAs, the generator is the seed
of the sequence. For such a generated (infinite) group π, we compute the SL(2,C) character variety,
where SL(2,C) is simultaneously a ‘space-time’ (a Lorentz group) and a ‘quantum’ (a spin) group.
A noteworthy result of our approach is to recognize that optimal regulation occurs when π looks
similar to a free group Fr (r = 1 to 3) in the cardinality sequence of its subgroups, a result obtained in
our previous papers. A non-free group structure features a potential disease. A second noteworthy
result is about the structure of the Groebner basis G of the variety. A surface with simple singularities
(such as the well known Cayley cubic) within G is a signature of a potential disease even when π
looks similar to a free group Fr in its structure of subgroups. Our methods apply to groups with a
generating sequence made of two to four distinct DNA/RNA bases in {A, T/U, G, C}. We produce a
few tables of human TFs and miRNAs showing that a disease may occur when either π is away from
a free group or G contains surfaces with isolated singularities.
Keywords: transcription factors; microRNAs; diseases; finitely generated group; SL(2,C) character
variety; algebraic surfaces
1. Introduction
Again since the one face, constant in symmetry, appears sometimes fair and
sometimes not, can we doubt that beauty is something more than symmetry, that
symmetry itself owes its beauty to a remoter principle?
[1] (Ennead I, Sixt Tractate, p66).
The remote principle envisaged by Plotinus is still a symmetry principle but in a
modern definition involving group theory and algebraic geometry. Recently, we wrote a
paper about a common algebra possibly ruling the beauty and structure in poems, music
and proteins [2]. We found that free groups govern the structure of such disparate topics
where a language emerges from pure randomness. We coined the concept of ‘syntactical
freedom’ for qualifying the occurrence of symbols and rules organized according to group
theory and aperiodicity [3,4]. One favorite decomposition of the secondary structure of
proteins is in term of α-helices, β-sheets and coils [5] but this decomposition and the
resulting syntax is model dependent [2]. According to our view, at the genome scale,
the escape to ‘syntactical freedom’ means a lack of harmony and the signature of a potential
disease. The secondary structures in the sequence of proteins or viruses are, most of the
time, organized according to the rules of free groups and, otherwise they may be a witness
of a potential aberrant topology.
Symmetry 2023, 15, 770. https://doi.org/10.3390/sym15030770
https://www.mdpi.com/journal/symmetry
Symmetry 2023, 15, 770
2 of 14
Of course, there are other techniques to identify a potential disease in the DNA/RNA
structures. The loose of homochirality of DNA may indicate an age-related disease [6]. Data
mining techniques may be employed to identify cancer in gene expression [7]. The role of
miRNAs in the regulation of pathogenesis during infection is investigated in [8].
Apart from the canonical double helix B-DNA we now know that there exists a
diversity of non-canonical coding/decoding sequences organized in structures such as
Z-DNA (often encountered in transcription factors [9,10]), G-quadruplex (in telomeres) and
other types that are single-stranded, two-stranded, or multistranded [11]. RNA is usually
a single-stranded molecule in a short chain of nucleotides, as is the case for a messenger
RNA (mRNA) or a (non-coding) microRNA (miRNA).
In our approach, the investigated sequences define a finitely generated group fp
whose structure of subgroups is close or away from a free group Fr of rank r, where r + 1 is
the number of distinct amino acids in the sequence (or the number of distinct secondary
structures considered in the protein chain) [4]. Recently, we also introduced concepts for
representing the groups fp over the Lie group SL(2,C). The SL(2,C) character variety
of fp and its Groebner basis are topological ingredients, they feature algebraic geometric
properties of the group fp under question [12–14]. A reminder of the theory is given in the
next section.
In this paper, we focus upon transcription factors and miRNAs, both serve at properly
decoding and regulating the genes and their action, either independently of each other or
together by targeting common genes [15]. Figure 1 (Left) is a picture of the pluripotent
transcription factor Nanog. Figure 1 (Right) is an example of a pre-miRNA associated to a
disease [16]. Both are investigated in detail in this paper.
Figure 1. Left: the Nanog transcription factor (PDB 9ANT). Right: the pre-miR-155 secondary
structure [16].
In Section 2, we describe the mathematical methods and the software needed for
describing the algebraic surfaces relevant to DNA/RNA sequences. This includes the
definition of infinite groups under question, of the free groups Fr of rank r = 1 to 3 corre-
sponding to 2- to 4-base sequences and the calculation of SL(2,C) representation of such
groups. A special care is needed to compute a Groebner basis of the character variety.
Symmetry 2023, 15, 770
3 of 14
Section 3 is a discussion about the type of singular surfaces encountered in this research.
They play an important role in our view of the discrimination of a potential disease.
In Section 4, the methods are applied to representative examples of sequences taken
from transcription factors and microRNAs whose group is close or away from a free group,
and whose Groebner basis of the variety contains simple singularities. Our theory applies
to many TFs and miRNAs that are known to be related to an identified disease.
Section 5 summarizes our paper and opens a few perspectives.
2. Theory
2.1. Finitely Generated Groups, Free Groups and Their Conjugacy Classes
Let Fr be the free group on r generators (r is called the rank).
The number of conjugacy classes of Fr of a given index d is known and is a signature
of the isomorphism, or the closeness, of a group π to Fr [4,17]. The cardinality structure of
conjugacy classes of index d in Fr will be called the card seq of Fr. We need the cases from
r = 1 to 3 that correspond to the number of distinct bases in a DNA/RNA sequence. The card
seq of Fr is in Table 1 for the 3 sequences of interest in the context of DNA/RNA.
The free group F1 of rank 1 may be defined as F1 = 〈a|∅〉 (with one generator a and
no relation) or F1 = 〈a, b|ab〉 (with two generators a and b and the relation ab). Similarly,
the free group F2 of rank 2 is F2 = 〈a, b|∅〉 = 〈a, b, c|abc〉 and so on for higher rank r.
Table 1. The counting of conjugacy classes of subgroups of index d in the free group Fr of rank r = 1 to
3. The last column is the index of the sequence in the on-line encyclopedia of integer sequences [18].
r
Card Seq
Sequence Code
1
[1, 1, 1, 1, 1, 1, 1, 1, 1, · · · ]
A000012
2
[1, 3, 7, 26, 97, 624, 4163, 34470, 314493, · · · ]
A057005
3
[1, 7, 41, 604, 13753, 504243, 24824785, 1598346352, · · · ]
A057006
Next, given a finitely generated group f p with a relation (rel) given by a sequence
motif, we are interested in the number of conjugacy classes of subgroups of index d (the
card seq of fp). Often, for a selected DNA/RNA motif taken as the generator of a finitely
generated group fp, the card seq that is obtained is close to that of a free group Fr, with r+ 1
being the number of distinct bases involved in the motif.
The closeness of fp to Fr can be checked by its signature in the finite range of indices
of the card seq.
To illustrate our methodology, let us consider the seed UCCUACA of the microRNA
sequence mir-155-3p that is investigated in detail in Section 4.3; see also Figure 2. The finitely
generated group is fp = 〈U, A, C|UCCUACA〉. The card seq of fp is found to be the series
[1, 3, 10, 51, 164, 1230 · · · ]. The group fp is of rank 2 and the card seq corresponds to that
of the group we call π2 in the tables of Section 4. In this case, the card seq of fp is ‘away’
from the card seq of the free group of the same rank F2 and the two groups fp and F2 are of
course ‘away’ to each other. However, if we omit the last nucleotide A of the seed in the
generator of fp, then the group fp obtains the same card seq than F2 (at least in the finite
range of the card seq series that we can check) and, in this sense, both groups fp and F2
become ‘close’ to each other.
2.2. The SL(2,C) Character Variety of a Finitely Generated Group and a Groebner Basis
Let fp be a finitely generated group, we describe the representations of fp in SL(2,C),
the group of (2× 2) matrices with complex entries and determinant 1. The group SL(2,C)
may be seen simultaneously as a ‘space-time’ (a Lorentz group) and a ‘quantum’ (a spin) group.
Such a group describes representations as degrees of freedom for all quantum fields
and is the gauge group for Einstein–Cartan theory. The later contains the Einstein–Hilbert
action and Einstein’s field equations [19]. The so-called Holst action used in loop quantum
gravity has quantum gravity states given in terms of SL(2,C) representations [20].
Symmetry 2023, 15, 770
4 of 14
Figure 2. (Up): Complementary base-pairing between miR-155-3p and the human Irak3 (interleukin-1
receptor-associated kinase 3) mRNA ([16], Figure 5). The requisite‘seed sequence’ base-pairing is
denoted by the bold dashes. (Down): the surface f (A1)
b
(x, y, z) = x2 + y2 − 6z2 + 4xyz.
Representations of fp in SL(2,C) are homomorphisms ρ : fp → SL(2,C) with charac-
ter κρ(g) = tr(ρ(g)), g ∈ fp. The notation tr(ρ(g)) means the trace of the matrix ρ(g). The
set of characters allows to determine an algebraic set by taking the quotient of the set of
representations ρ by the group SL(2,C), that acts by conjugation on representations [21,22].
Such an algebraic set is called the SL(2,C) character variety of fp. The set is made of a
sequence of multivariate polynomials called a scheme X. The ideal I(X) is defined by the
vanishing of polynomials living in the scheme X. Below, we use a particular basis G(X) of
the polynomial ring I(X), called a Groebner basis. The Groebner basis G(X) has to follow
algorithmic rules (similar to the Euclidean division for univariate polynomials) [14].
For the effective calculations of the character variety, we make use of a software on
Sage [23]. We also need Magma [24] for the calculation of a Groebner basis, at least for 3-
and 4-base sequences.
2.3. Algebraic Geometry and Topology of DNA/RNA Sequences
2.3.1. Two-Base Sequences
Following [25], in this section, we are interested by the special case of representations
for the once punctured torus S1,1 and the relevance of the extended mapping class group
Mod±(S1,1) in its action on surfaces of type κd(x, y, z), d ∈ C.
The fundamental group of T1,1, that we denote π, is the free group F2 = 〈a, b|∅〉 on
two generators a and b. The boundary component of T1,1 consists of a single loop around
the puncture expressed by the commutator [a, b] = abAB with A = a−1 and B = b−1.
Taking the traces x = tr(ρ(a)), y = tr(ρ(b)), z = tr(ρ(ab)), the trace of the commutator is
the surface [21,25]
tr([a, b]) = κ2(x, y, z) = x2 + y2 + z2 − xyz− 2.
According to the Dehn–Nielsen–Baer theorem [26], for a surface of genus g ≥ 1,
we have
Mod±(Sg) ∼= Out(π(Sg)),
where Mod(S) is the mapping class group. It denotes the group of isotopy classes of
orientation-preserving diffeomorphisms of S.
Symmetry 2023, 15, 770
5 of 14
The extended mapping class group Mod±(S) is the group of isotopy classes of all
homeomorphisms of S (including the orientation-reversing ones). The outer automorphism
group of π is denoted Out(π). This leads to the (topological) action of Mod± on the once
punctured torus as follows
Mod±(S1,1) = Out(F2) = GL(2,Z).
(1)
The automorphism group Aut(F2) acts by composition on the representations ρ and
induces an action of the extended mapping class group Mod± on the character variety by
polynomial diffeomorphisms of the surface κd defined by [25]
f (4)
H (x, y, z) = κd(x, y, z) = xyz− x
2 − y2 − z2 + d.
(2)
Within the family κd(x, y, z), the Cayley surface κ4(x, y, z) is obtained from the char-
acter variety for the fundamental group of the Hopf link L2a1 (the link of two unknotted
curves). For the Hopf link, the fundamental group is
π(S3 \ L2a1) = 〈a, b|[a, b]〉 = Z2,
where the notation S3 is for the 3-sphere and \L2a1 means that a small tubular neighbor-
hood around the Hopf link L2a1 is removed from S3 to define the corresponding Hopf
link 3-manifold.
The Cayley surface κ4(x, y, z) possesses 4 simple (isolated) singularities. At these
points, the surface loses its smoothness. We already showed that it plays a role in the
context of Z-DNA conformations of transcription factors ([13], Tables 2 and 5); see also,
Section 4 below, and notably Figure 3 (Left).
The surface κ3(x, y, z) lies within the character variety for the fundamental group of
the link L6a1 [27]. We show below that this surface also lies in the generic Groebner basis
obtained for 4-base sequences; see Section 4 below.
Figure 3. (Left): the Cayley cubic κ4(x, y, z). (Right): the surface f
(A1)
a
(x, y, z).
2.3.2. Three-Base Sequences
Our main object in this section is the four punctured sphere for which the fundamental
group is the free group F3 of rank 3 whose character variety generalizes the Fricke cubic
surface (2) to the surface Va,b,c,d(C) in C7. From now, a and b are no longer generators of a
group but belong to the set of parameters of the surface V.
We follow the work of references [21,25,28].
Let S4,2 be the quadruply punctured sphere. The fundamental group for S4,2 can be
expressed in terms of the boundary components x1, x2, x3, x4 as
π(S4,2) = 〈x1, x2, x3, x4|x1x2x3x4〉 ∼= F3.
Symmetry 2023, 15, 770
6 of 14
A representation π → SL(2,C) is a quadruple
α = ρ(A), β = ρ(B) γ = ρ(C), δ = ρ(D) ∈ SL(2,C) where αβγδ = I.
Let us associate the seven traces
a = tr(ρ(α)), b = tr(ρ(β)), c = tr(ρ(γ)), d = tr(ρ(δ))
x = tr(ρ(αβ)), y = tr(ρ(βγ)), z = tr(ρ(γα)),
where a, b, c, d are boundary traces and x, y and z are traces of elements AB, BC and CA
representing simple loops on S4,2.
The character variety for S4,2 satisfies the equation ([21], Section 5.2), ([25], Section 2.1),
([28], Section 3B), ([29], Equation (1.9)) or ([30], Equation (39)).
Va,b,c,d(x, y, z) = x2 + y2 + z2 + xyz− θ1x− θ2y− θ3z− θ4 = 0
(3)
with θ1 = ab+ cd, θ2 = ad+ bc, θ3 = ac+ bd and θ4 = 4− a2 − b2 − c2 − d2 − abcd.
The 4-punctured sphere, whose fundamental group is the free group F3 with generator
the product of the 4 letters, is a generic topology. It is straightforward to check that the
Groebner basis for F3 contains (among other surfaces and depending on the choice of
parameters) a single copy of the generic surfaces κ4(x, y, z), κ3(x, y, z) and V1,1,1,1(x, y, z) =
xyz + x2 + y2 + z2 − 2x − 2y− 2z + 1, a surface we also denote f (3A1)(x, y, z) because it
contains 3 simple singularities of type A1 as shown in Figures 4 and 5.
There are other surfaces encountered in our study of the Groebner basis for transcrip-
tion factors and miRNAs when the generated group is close or away from the free group
F2 (for 3-base sequences) or the free group F3 (for 4 base sequences). These surfaces are
described in Section 4.
Figure 4. The Fricke surface V1,1,1,1(x, y, z) = f
(3A1)
a
(x, y, z) (with three simple singularities of type A1).
2.3.3. Four-Base Sequences
There does not exist a huge difference in the structure of a Groebner basis of the
character variety in the case of a 4-base sequence compared to the case of a 3-basis
sequence. One difference is that one has to manage a 14-dimensional hypersurface
Va,b,c,d,e, f ,g,h(x, y, z, u, v, w) in C14 (instead of a 7-dimensional one as in the previous sub-
section). In general, after the appropriate choice of the 8 parameters a, b, c, d, e, f , g, h,
the Groebner basis contains more than one copy of the generic Groebner basis. Each copy S
of a relevant surface may be of the form S(x, y, z), S(x, u, v), S(y, u, w) or S(z, v, w).
Symmetry 2023, 15, 770
7 of 14
Figure 5. (Up): Complementary base-pairing between miR-155-5p and the human Spi1 (spleen focus
forming virus proviral integration oncogene) ([16], Figure 4). The requisite ‘seed sequence’ base-
pairing is denoted by the bold dashes. (Down (from left to right)): the surfaces f (4)
H = κ4(x, y, z),
f (3A1)(x, y, z) and κ3(x, y, z), four copies of them are contained within the Groebner basis for the
character variety.
3. Discussion
Given an ordinary projective surface S in the projective space P3 over a number field,
when S is birationally equivalent to a rational surface, the software Magma [24] determines
the map to such a rational surface and returns its type within five categories. The returned
type of S is P2 for the projective plane, a quadric surface (for a degree 2 surface in P3),
a rational ruled surface, a conic bundle or a degree p Del Pezzo surface where 1 ≤ p ≤ 9.
One important attribute of a projective surface is its degree of singularity. Most surfaces
S of interest below are almost not singular in the sense that they have at worst simple
(isolated) singularities. A simple singularity is referred to as an A-D-E singularity [31]. It
has to be of the type An, n ≥ 1, Dn, n ≥ 4, E6, E7 or E8.
The type and the number of simple singularities are denoted in an exponent such
as S(lA1) for l singularities of type A1. All such surfaces are degree 3 del Pezzo surfaces.
For the Cayley cubic f (4)(x, y, z), the exponent (4) means (4A1).
There are additional facilities offered by Magma for studying a singular scheme.
As already mentioned, a scheme is any geometric object defined by a set of polynomials in
a projective space. An algebraic surface is a scheme. If the scheme has simple singularities,
one can calculate the degree and the support of the reduced singular subscheme that are
signatures of the scheme. In the examples of this paper, for a surface S(lA1), the degree is
l and the support contains l simple singular points. Otherwise, we add a lower index to
S(lA1) to qualify the index and the support of the singular subscheme. The notation S(lA1)
m,{}
means that there are l singularities of type A1, that the degree of the singular subscheme is
m and that the support is the empty set.
Most of the time, the surfaces in the character variety attached to a transcription factor
or a microRNA only contain isolated singularities.
This contrasts with some sequences encountered in another context. For instance,
a complete turn of A-DNA (PDB; 2D47) defines the dodecamer sequence CCCCCGCGGGGG
whose attached character variety contains the surface ([13], Figure 5).
fH̃(x, y, z) = z
4 − 2xyz+ 2x2 + 2y2 − 3z2 − 4.
(4)
Symmetry 2023, 15, 770
8 of 14
This surface contains many non isolated singularities (that cannot be resolved by blow
ups). Methods to desingularize a surface of such a type are presented in a companion
paper [32].
To summarize the important issues below, a noteworthy result of our approach is to
recognize that optimal regulation occurs when the group underlying the sequence looks
similar to a free group Fr (r = 1 to 3) in the cardinality sequence of its subgroups, a result
obtained in our previous papers. A non-free group structure features a potential disease.
A second noteworthy result is about the structure of the Groebner basis of the variety.
A surface with simple singularities (such as the well known Cayley cubic) within the
Groebner basis is a signature of a potential disease even when the generated group looks
similar to a free group Fr in its structure of subgroups. Our methods apply to groups with
a generating sequence made of two to four distinct DNA/RNA bases in {A, T/U, G, C}.
Several human TFs and miRNAs are investigated in detail thanks to our approach. We
summarize this discussion in the diagram presented in Figure 6.
Figure 6. A diagram with the main results discussed in the main text.
4. Results
In this section, we apply the SL(2,C) representation theory to groups generated by
DNA/RNA sequences occurring in transcription factors and microRNAs. Both play a
leading role in the decoding of the genome and in genome-scale regulatory networks.
Two-letter transcription factors (TFs) whose structure is close or away from the free group
F1 were already investigated in ([13], Table 2). The occurrence of the Cayley cubic κ4(x, y, z)
in the Groebner basis of the character variety was found to be a signature in the former case.
In this case, this surface seems to possess a regulatory action that may be lost in the latter
case. In ([13], Table 3), the potential diseases associated with a non-free group structure are
mentioned. In the present section, one explores, in Table 2, the case of a 3-letter sequence of
a TF and in Table 3, the case of 3-letter sequence of a miRNA. The case of a 4-letter sequence
of a miRNA is summarized in Table 4. The role of surfaces with simple singularities in the
Groebner basis is emphasized.
4.1. Algebraic Morphology of the Transcription Factor Prdm1
The transcriptional repressor PR domain containing 1 (Prdm1), also known as B-
lymphocyte-induced maturation protein-1 (Blimp1), is essential for normal development
and immunity [33]. It is of a zinc finger type. The consensus sequence ACTTTC corresponds
to the code MA0508.2 in [34].
Symmetry 2023, 15, 770
9 of 14
4.1.1. The Character Variety
The ideal for the character variety fPrdm1(a, b, c, d)(x, y, z) for a few values of the
parameters is
fPrdm1(0, 0, 0, 0) = κ−4(x, y, z)(yz+ x + 2),
fPrdm1(0, 1, 1, 0) = yκ−2(x, y, z)(x− 1),
fPrdm1(0, 1, 0, 0) = zκ−3(x, y, z)(z
2 + 1)(yz+ x + 1)(yz+ x + 2),
fPrdm1(1, 1, 1, 1) = f
(3A1)
a
(x, y, z)(y+ 1)(y+ z− 1),
where κ−2(x, y, z), κ−3(x, y, z) are Fricke surfaces [27] and f
(3A1)
a
(x, y, z) = xyz+ x2 + y2 +
z2 − 2x− 2y− 2z+ 1 is the surface drawn in Figure 4. The subscript 3A1 is for featuring
the three singularities of type A1.
4.1.2. The Groebner Basis
The singular surfaces found in the Groebner basis of the ideal are not similar to those
in the ideal. One of them S1 = S
(A1A3)
3,{0:1:0:0} = 2yz
2 + x2 + 3z2 − 2xz− 2yz− 2y− 2z− 2 is
obtained at values (a, b, c, d) = (1, 1, 1, 1) featuring two simple singularities of type A1 and
A3 with a singular subscheme of degree 3 and the singular point of type A3 in its support.
The other surface obtained in the Groebner basis at values (a, b, c, d) = (0, 0, 0, 0) is a conic
bundle of the K3 type S2 := z4 + 2yz3 + x2 − 6yz− 2x− 8 whose singular subscheme is
non zero dimensional and of degree 1.
These two surfaces are non standard in our context of TFs and miRNAs. The formal
desingularization of the surface S2 is given in [32].
4.2. Algebraic Morphology of Homeodomains for Nanog and Xvent
The pluripotency in embryonic stem cells and their regulation is characterized by the
expression of several transcription factors [35,36]. Among them, the transcription factor
Nanog is present in the embryonic stages of life of several vertebrate species. Nanog binds
to promoter elements of hundreds of target genes as a regulatory element. It has a conserved
DNA-binding homeodomain with consensus sequence TAATGG. The closest homolog
of Nanog is the (nonmammalia) Xenopus, a Xvent transcription factor with consensus
sequence CTAATT [36]. In this subsection, we investigate the algebraic morphology of both
transcription factors Nanog and Xvent thanks to their consensus sequences.
Table 2. A few (three-base) transcription factors whose group structure is away from a free group or
whose Groebner basis of the SL(2,C) character variety contains a (possibly almost) singular surface.
The symbol gene is for the identification of the transcription factor in the Jaspar database [34], motif
is for the consensus sequence of the transcription factor, card seq is for the cardinality sequence of
conjugacy classes of subgroups of the group whose motif is the generator, simple sing is for the
identification of a surface with simple singularities within the Groebner basis and the last column
is for a reference paper and the corresponding disease. The group F2 is the free group of rank
two. The card seq for π2 is [1, 3, 10, 51, 164, 1230, 7829, 59835, 491145 · · · ], close to the card seq of
the group 〈x, y, z|(x, (y, z)) = z〉. The latter group is found as governing the structure of many
transcription factors and is associated to the link found in ([13], Figure 2). The card seq for π3 is
[7, 14, 89, 264, 1987, 11086, 93086 · · · ]. The surface f (A1)
b
(x, y, z) = x2 + y2 − 6z2 + 4xyz (not defined in
the text) is part of the character variety for the genes Pitx1, OTX1, etc.
Gene
Motif
Card Seq
Simple Sing
Ref & Disease
Prdm1
ACTTTC
F2
S1, S2(x, y, z)
[34], MA0508.2
lupus, rheumatoid arthritis
Symmetry 2023, 15, 770
10 of 14
Table 2. Cont.
Gene
Motif
Card Seq
Simple Sing
Ref & Disease
POU6F1
TAATGAG
π2
no
MA1549.1
lung adenocarcinoma
ELK4
CTTCCGG
.
no , Fricke
MA0076.2
gastric cancer
OTX2
GGATTA
π3
no
[MA0712.2, MA0883.1]
medulloblastomas
N-box
TTCCGG
.
no, Fricke
[37]
drug sensitivity
Pitx1,OTX1,· · ·
TAATCC
.
f (4)
H
, f (A1)
b
(x, y, z)
[34], [MA0682.1,MA0711.1]
autism, epilepsy, · · ·
Nanog
TAATGG
.
f (4)
H
, f (A1)
a
(x, y, z)
[35]
cancer cells
Xvent
CTAATT
F2
f (2A1)
4,{}
, f (A2)(x, y, z)
[36]
The Groebner basis for Xvent fNanog(0, 0, 0, 0) takes the form
fNanog(0, 0, 0, 0) = f
(4)
H (x, y, z) f
(A1)
a
(x, y, z) · · ·
where f (4)
H (x, y, z) is the Cayley cubic (with its 4 simple singularities) and f
(A1)
a
(x, y, z) =
x2 + y2 − z2 + xyz (a surface with a single simple singularity of type A1) as shown in
Figure 3 (Right). The forgotten factors are factors for planes or trivial smooth surfaces.
The Groebner basis for Xvent fXvent(1, 1, 1, 1) takes the form
fXvent(1, 1, 1, 1) = f
(3A1)
b
(x, y, z) · · ·
where f (3A1)
b
(x, y, z) = x2 + y2 + xyz− xy− z− 1 (a surface with three simple singularity
of type A1). The missing term does not contain surfaces with singularities. The character
variety fXvent(0, 0, 0, 0) contains the cubic surface f
(2A1)
4,{} (x, y, z) = 2z
3 + x2z + 2xyz +
2y2− z2− 6z (with two simple singularities of type A1) and other factors for planes or trivial
smooth surfaces. Both surfaces f (3A1)
b
(x, y, z) and f (2A1)
4,{} (x, y, z) are pictured in Figure 7.
Figure 7. (Left): the cubic surface f (2A1)
4,{} (x, y, z). (Right): the cubic surface f
(3A1)
b
(x, y, z).
Table 2 lists a few selected transcription factors, their card seq and the corresponding
singular surfaces, if any. As announced, in the selected transcription factors, there exists
a correlation between the lack of ‘syntactical freedom’, or the presence of a surface with
isolated singularities in the character variety, with an identified disease.
Symmetry 2023, 15, 770
11 of 14
4.3. Algebraic Morphology of microRNAs
MicroRNAs (miRNAs) are important players of the expression and regulation of
genes by targeting specific messenger RNAs (mRNAs) for degradation or translational
repression [38,39]. The miRNAs are approximately 22 nt (where nt is for nucleotide) long
single-stranded RNA molecules. The genes encoding miRNAs have a length much longer
than the processed mature miRNA molecule. Many miRNAs reside in introns of their
pre-mRNA host genes. They share their regulatory elements, primary transcript and have a
similar expression profile. MicroRNAs are transcribed by RNA polymerase II as large RNA
precursors called pri-miRNAs. The pre-miRNAs are approximately 70-nucleotides in length
and are folded into imperfect stem-loop structures; see Figure 1 (Right) for an example.
Each miRNA is synthesized as a miRNA duplex comprising two strands (-5p and -3p).
However, only one of the two strands is selectively incorporated into the RNA-induced
silencing complex to act as a template for the transcript of a complementary mRNA [40,41].
For details about the mirRNA sequences, we use the Mir database [42,43].
Plant miRNAs usually have near-perfect pairing with their mRNA targets so that
gene repression proceeds through cleavage of the target transcripts. In contrast, animal
miRNAs are able to recognize their target mRNAs by using as few as 6 to 8 nucleotides
(the seed region), which is not enough pairing for leading to cleavage of the target mRNAs.
A given miRNA may have hundreds of different mRNA targets, and a given target might
be regulated by multiple miRNAs.
Disregulation of miRNAs may lead to a disease such as cancer. A key microRNA
known as an oncommir (involved in immunity and cancer) is mir-155.
Specifically the -3p strand is mir-155-3p. Figure 2 (top) illustrates the complementary
base-pairing between miR-155-3p and the human IRAK3 (interleukin-1 receptor-associated
kinase 3) mRNA ([16], Figure 4) and the relevant seed sequence UCCUAC. The card seq for
this sequence is the two-letter free group F2 and the Groebner basis for the corresponding
character variety contains the surface f (A1)
b
(x, y, z) = x2 + y2 − 6z2 + 4xyz that has a single
simple singularity as shown in Figure 2 (down). If one retains the full seed sequence is
UCCUAC(A) then the card seq passes to that of the free group F2 to the group π2 and the
singular surface is lost. This is a case where the ‘bandwidth’ of the seed is critical in the
(dis)regulation of the miRNA. These results are transcribed in Table 3.
Table 3. A few human (prefix ‘hsa’) microRNAs whose group structure is away from a free group
or whose Groebner basis of the SL(2,C) character variety contains a singular surface. The symbol
mir is for the identification in the Mir database [43], seed is for the seed of the miRNA, card seq
is for the cardinality sequence of conjugacy classes of subgroups of the group whose seed is the
generator, sing is the identification of a singular surface within the Groebner basis and the last column
is for a reference paper and the corresponding disease [40]. The card seq for π1 and π′1 are given in
([4], Table 5). The card seq for π′2 is [1, 3, 7, 34, 139, 931, 5208, 43867 · · · ]. For hsa-mir-124-1-3p, one
encounters the Fricke surface f (A1)
2,{} = xyz+ x
2 + y2 + z2 − 2y in the character variety.
mir
Seed
Card Seq
Simple Sing
Ref & Disease
hsa-mir-193b-5p
GGGGUU
π1
no
[40,43]
GGGGUUU
π′1
no
lung cancer
hsa-mir-155-3p
UCCUAC
F2
f (A1)
b
(x, y, z)
[40,41,43]
UCCUACA
π2
no
multiple sclerosis
hsa-mir-193a-5p GGGUCUU
F2
f (A1)
b
(x, y, z)
[40,43]
breast cancer
hsa-mir-223-5p
GUGUAUU
.
.
.
hsa-mir-133-3p
UUGGUC
F2
f (3A1)
b
(x, y, z)
[40,43]
UUGGUCC
π′2
no
atrial fibrillation
Symmetry 2023, 15, 770
12 of 14
Table 3. Cont.
mir
Seed
Card Seq
Simple Sing
Ref & Disease
hsa-mir-124-3p
AAGGCA
F2
f (3A1)
b
, f (A1)
2,{}
[43,44]
AAGGCAC
.
no sing
Alzheimer’s disease
For the case of -5p strand mir-155-3p, the seed sequence UUAAUGCUA contains
four distinct letters. This case is similar to generic Groebner bases obtained from four
letter seeds. Depending on the choices of parameters a, b, c, d, e, f , g, h, the Groebner basis
contains the Cayley cubic f (4)
H (x, y, z), the Fricke surface κ3(x, y, z) (that is related to the
link L6a1 ([27], Figure 2)), the surface f (3A1)
a
(x, y, z) shown in Figure 4 and other surfaces.
In this generic case, the surface is found with (at most) 4 copies where each copy is attached
to a distinct puncture of the 4-punctured 4-sphere S4,2.
In Table 3, this generic case is denoted 4× generic (or 3× generic for mir-133-5p).
These results are transcribed in Table 4.
Table 4. The opposite strand of the microRNA considered in Table 3. The seed sequence is made
of 4 distinct bases and the corresponding card seq is the free group F3 of rank 3. The Groebner
basis contains 4 copies of the generic collection of surfaces κ4(x, y, z), f (3A1)(x, y, z), κ3(x, y, z), etc.,
as shown in Figure 5, except for the -5p strand of mir-133, where there are only 3 copies of the
generic surfaces.
mir
Seed
Card Seq
Sing
Ref & Disease
hsa-mir-193b-3p
ACUGGCC
F3
4× generic
[40,43]
hsa-mir-155-5p
UUAAUGCUA
.
.
[40,41,43]
hsa-mir-193a-3p
ACUGGCC
.
.
[40,43]
hsa-mir-223-3p
GUCAGUU
.
.
.
hsa-mir-124-5p
GUGUUCA
.
.
.
hsa-mir-133-5p
GCUGGUA
.
3× generic
[43,44]
A small list of huma miRNAs is investigated in Tables 3 and 4 corresponding to 3-letter
and 4-letter seeds. the prefix ‘hsa’ is for the human species. Similar to transcription factors
in Table 2, the lack of ‘syntactical freedom’, or the occurrence of a singular surface in the
character variety, is symptomatic of a disease.
5. Conclusions
We found, in this work, that a signature of a disease may be given in terms of the
group structure of a DNA/RNA sequence and the related character variety representing
the group. The DNA motif of a transcription factor, or the seed of a microRNA, defines the
generator of a group π. As soon as π is away from a free group Fr (with r + 1 the number
of distinct bases in the sequence) or the SL(2,C) character variety G of π contains singular
surfaces with isolated singularities, a potential disease is on sight, as shown in Figure 6.
For example, for mir-155-3p, examined in much detail in Section 4.3, the seed UCCUAC
serves as the generator of the group π whose card seq is that of the free group F2 and
whose Groebner basis of the variety contains the surface f (A1)
b
(x, y, z) possessing an isolated
singularity of type A1. The potential disease is multiple sclerosis. Note that a longer seed
(with A added at the right hand side) is not appropriate to the detection. A partial list
of other potential diseases identified from the structure of a miRNA seed can be seen
in Table 3. The diseases are a lung cancer associated with mir-193b-5p, a breast cancer
associated with mir-194a-5p or mir-223-5p, an atrial fibrillation associated with mir-133-3p
and an Alzheimer’s disease associated with mir-124-3p.
One would like to be more predictive in identifying the potential disease with peculiar
groups or singular surfaces. First of all, most of the time, the surfaces encountered in the
context of TFs and miRNAs are degree 3 del Pezzo, in contrast to surfaces obtained from
Symmetry 2023, 15, 770
13 of 14
other DNA sequences, as in Equation (4). However, the degree 3 del Pezzo family is very
rich. For instance, the singular surface f (A1)
b = x
2 + y2 − 6z2 + 4xyz (see Figure 2) is part
of the character variety of TF Pitx1 (see Table 2) and of miRNAs 155-3p and 193a-5p (in
Table 3). Then, the singular surface f (A1)
2,{} = x
2 + y2 + z2 + xyz− 2y is part of the character
variety of mirRNA 133-3p. Both surfaces have a simple singular point of type A1 but
distinct singular subschemes (see Section 3 for the notation).
An exception to the degree 3 del Pezzo rule was found in investigating the character
variety for the Prdm1 transcription factor in Section 4.1. Do these features and other ones
to be described later help for the diagnostic of a potential disease? There is room for much
work in the future along these lines, as shown in our recent new paper [32].
In a separate direction, the work of Reference [45] about biochirality and the CPT
theorem may be possibly related to the space-time-spin group SL(2,C) that served as a
prism of DNA/RNA structure in our approach. Related ideas about the concept of a time
crystal are also worthwhile to be pointed out, e.g., Reference [46].
Author Contributions: Conceptualization, M.P.; methodology, M.P and M.M.A.; software, M.P.;
validation, M.M.A. and K.I.; formal analysis, M.P.; investigation, M.M.A.; resources, K.I.; data
curation, M.P.; writing—original draft preparation, M.P.; writing—review and editing, M.P. and
M.M.A.; visualization, M.M.A.; supervision, M.P.; project administration, K.I.; funding acquisition,
K.I. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Informed Consent Statement: Not applicable.
Data Availability Statement: Data are available from the authors after a reasonable demand.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Uz̆davinys, A. The Heart of Plotinus. The Essential Enneads; World Wisdom, Inc.: Bloomington, IN, USA, 2009.
2.
Planat, M.; Aschheim, R.; Amaral, M.M.; Fang, F.; Irwin, K. Graph coverings for investigating non local structures in protein,
music and poems. Science 2021, 3, 39. [CrossRef]
3.
Irwin, K. The code-theoretic axiom; the third ontology. Rep. Adv. Phys. Sci. 2019, 3, 39. [CrossRef]
4.
Planat, M.; Amaral, M.M.; Fang, F.; Chester, D.; Aschheim, R.; Irwin, K. Group theory of syntactical freedom in DNA transcription
and genome decoding. Curr. Issues Mol. Biol. 2022, 44, 1417–1433. [CrossRef] [PubMed]
5.
Dang, Y.; Gao, J.; Wang, J.; Heffernan, R.; Hanson, J.; Paliwal, K.; Zhou, Y. Sixty-five years of the long march in protein secondary
structure prediction: The final strech? Brief. Bioinform. 2018, 19, 482–494.
6.
Dyakin, V.V.; Wisniewski, T.M.; Lajtha, A. Racemization in Post-Translational Modifications Relevance to Protein Aging, Aggrega-
tion and Neurodegeneration: Tip of the Iceberg. Symmetry 2021, 13, 455. [CrossRef]
7.
Abd El Nabi, M.L.R.; Jasim, M.W.; El-Bakry, H.M.; Taha, M.H.N.; Khalifa, N.E.M. Breast and colon cancer classification from gene
expression profiles using data mining techniques. Symmetry 2020, 12, 408. [CrossRef]
8.
Milhem, Z.; Chroi, P.; Nutu, A.; Ilea, M.; Lipse, M.; Zanoaga, O.; Berindan-Neagoe, I. Non-coding RNAs and reactive oxygen
species—Symmetric players of the pathogenesis associated with bacterial and viral infections. Symmetry 2021, 13, 1307. [CrossRef]
9.
Aldrich, P.R.; Horsley, R.K.; Turcic, S.M. Symmetry in the language of gene expression: A survey of gene promoter networks in
multiple bacterial species and non-regulons. Symmetry 2011, 3, 750–766. [CrossRef]
10. Heinemann, U.; Roske, Y. Symmetry in nucleic-acid double helices. Symmetry 2020, 12, 737. [CrossRef]
11.
Bansal, A.; Kaushik, S.; Kukreti, S. Non-canonical DNA structures: Diversity and disease association. Front. Genet. 2022, 13, 959258.
[CrossRef]
12.
Planat, M.; Aschheim, R.; Amaral, M.M.; Fang, F.; Chester, D.; Irwin, K. Character varieties and algebraic surfaces for the topology
of quantum computing. Symmetry 2022, 14, 915. [CrossRef]
13.
Planat, M.; Amaral, M.M.; Fang, F.; Chester, D.; Aschheim, R.; Irwin, K. DNA sequence and structure under the prism of group
theory and algebraic surfaces. Int. J. Mol. Sci. 2022, 23, 13290. [CrossRef]
14. Gröbner Basis. Available online: https://en.wikipedia.org/wiki/Gröbner_basis (accessed on 1 August 2022).
15. Martinez, N.J.; Walhout, A.J.M. The interplay between transcription factors and microRNAs in genome-scale regulatory networks.
Bioessays 2009, 31, 435–445. [CrossRef]
16. miR-155. Available online: https://en.wikipedia.org/wiki/MiR-155 (accessed on 18 November 2022).
17. Kwak, J.H.; Nedela, R. Graphs and their coverings. Lect. Notes Ser. 2007, 17, 118.
18.
The On-Line Encyclopedia of Integer Sequences. Available online: https://oeis.org/book.html (accessed on 1 November 2022).
Symmetry 2023, 15, 770
14 of 14
19. Hehl, F.W.; von der Heyde, P.; Kerlick, G.D.; Nester, J.M. General relativity with spin and torsion: Foundations and prospects.
Rev. Mod. Phys. 1976, 48, 393–416. [CrossRef]
20. Rovelli, C.; Vidotto, F. Covariant Loop Quantum Gravity, 1st ed.; Cambridge University Press: Cambridge, MA, USA , 2014.
21. Goldman, W.M. Trace coordinates on Fricke spaces of some simple hyperbolic surfaces. Eur. Math. Soc. 2009, 13, 611–684.
22. Ashley, C.; Burelle J.P.; Lawton, S. Rank 1 character varieties of finitely presented groups. Geom. Dedicata 2018, 192, 1–19. [CrossRef]
23.
Python Code to Compute Character Varieties. Available online: http://math.gmu.edu/~slawton3/main.sagews (accessed on
1 May 2021).
24.
Bosma, W.; Cannon, J.J.; Fieker, C.; Steel, A. (Eds.) Handbook of Magma Functions, 2.23rd ed.; University of Sydney: Sydney,
Australia, 2017; 5914p.
25. Cantat, S. Bers and Hénon, Painlevé and Schrödinger. Duke Math. J. 2009, 149, 411–460. [CrossRef]
26.
Farb, B.; Margalit, D. A Primer on Mapping Class Groups; Princeton University Press: Princeton, NJ, USA, 2012.
27.
Planat, M.; Chester, D.; Amaral, M.; Irwin, K. Fricke topological qubits. Quant. Rep. 2022, 4, 523–532. [CrossRef]
28.
Benedetto, R.L.; Goldman W.M. The topology of the relative character varieties of a quadruply-punctured sphere. Exp. Math.
1999, 8, 85–103. [CrossRef]
29.
Iwasaki, K. An area-preserving action of the modular group on cubic surfaces and the Painlevé VI. Comm. Math. Phys. 2003, 242,
185–219. [CrossRef]
30.
Inaba, M.; Iwasaki, K.; Saito, M.H. Dynamics of the sixth Painlevé equation. arXiv 2005, arXiv:math.AG/0501007.
31. ADE Classification. Available online: https://en.wikipedia.org/wiki/ADE_classification (accessed on 1 August 2022).
32.
Planat, M.; Amaral, M.M.; Chester, D.; Irwin, K. SL(2,C) scheme processsing of singularities in quantum computing and genetics.
Axioms 2023, 12, 233. [CrossRef]
33. Doody, G.M.; Care, M.A.; Burgoyne, N.J.; Bradford, J.R.; Bota, M.; Bonifer, C.; Westhead, D.R.; Tooze, R.M. An extended set of
PRDM1/BLIMP1 target genes links binding motif type to dynamic repression. Nucl. Acids Res. 2010, 38, 5336–5350. [CrossRef]
[PubMed]
34.
Sandelin, A.; Alkema, W.; Engstrom, P.; Wasserman, W.W.; Lenhard, B. JASPAR: An open-access database for eukaryotic
transcription factor binding profiles. Nucleic Acids Res. 2004, 32, D91–D94. Available online: https://jaspar.genereg.net/ (accessed
on 1 November 2022). [CrossRef]
35.
Jauch, R. Crystal tructure and DNA inding of the homeodomain of the stem cell transcription factor Nanog. J. Mol. Biol. 2008, 376,
758–770. [CrossRef]
36.
Schuff, M.; Siegel, D.; Philipp, M.; Bunsschu, K.; Heymann, N.; Donow, C.; Knöchel, W. Characterization of Danio rerio Nanog
and Functional Comparison to Xenopus Vents. Stem Cells Devt. 2012, 21, 1225–1238. [CrossRef]
37.
Schaeffer, L.N.; Huchet-Dymanus, M.; Changeux, J.P. Implication of a multisubunit Ets-related transcription factor in synaptic
expression of the nicotinic acetylcholine receptor. EMBO J. 1998, 17, 3078–3090. [CrossRef]
38. microRNA. Available online: https://en.wikipedia.org/wiki/MicroRNA (accessed on 1 September 2022).
39.
Fang, Y.; Pan, X.; Shen, H.B. Recent deep learning methodology development for RNA-RNA interaction prediction. Symmetry
2022, 14, 1302. [CrossRef]
40. Medley, C.M.; Panzade, G.; Zinovyeva, A.Y. MicroRNA stran selection: Unwinding the rules. WIREs RNA 2021, 12, e1627.
[CrossRef]
41. Dawson, O.; Piccinini, A.M. miR-155-3p: Processing by-product or rising star in immunity and cancer? Open Biol. 2022, 12, 220070.
[CrossRef]
42. Kozomara, A.; Birgaonu, M.; Griffiths-Jones, S. miRBase: From microRNA sequences to function. Nucl. Acids Res. 2019, 47,
D155–D162. [CrossRef]
43. miRBase: The microRNA Database. Available online: https://www.mirbase.org/ (accessed on 1 November 2022).
44. Kou, X.; Chen, D.; Chen, N. The regulation of microRNAs in Alzheimer’s disease. Front. Neurol. 2020, 11, 288. [CrossRef]
45. Dyakin, V.V. Fundamental Cause of Bio-Chirality: Space-Time Symmetry—Concept Review. Symmetry 2023, 15, 79. [CrossRef]
46.
Sbitnev, V. Relativistic Fermion and Boson Fields: Bose-Einstein Condensate as a Time Crystal. Symmetry 2023, 15, 275. [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.