HOLOGRAPHIC ASSOCiATiVE MEMORY OF BiOLOGiCAL SYSTEMS

Peter P. Gariaev, Viktor I. Ohudin, and. Gennady G. Komissarov

Soientiic & Technical Center "MOSKVORECHIE",

36 Piatnitskaya Street, 109017 Moscow, USSR.
Andrey A. Berezin

Theoretical Problems Department, USSR Academy

o! Sciences, 12 Vesnina Street, Moscow, USSR.

Anatoly A. Vasiliev

P.N.Lebedev Physical Institute, The USSR Academy of

Sciences, 53 Leninsky Prospect, 117924, Moscow, USSR.

ABSTRACT
We consider some specific problems and phenomena of morphogenetic
infonnation storage, reproduction and transfer including phantom leaf ef-
feot and field—induced morphogenetic translations between different taxo—
nomic units. Several experimental results are presented and their expla—
nation is given using a new approach to morphogenesis which combines some
physical models or holographic associative memory and mathematical forma-
lism of Fermi — Pasta — Ulam recurrence for solitary waves in deoxyribo—
nucleic acid.

1 . INTRODUCTION

Some fundamental properties of eucaryote genome can not be reasonab—

include the genome "redundancy" , i

ly explained in the framework of conventional biology. These phenomena
5 mosaic s truc ture , spl io ing of pro—
7flRNA, mobile dispersive genes, coherent radiation of the chromosome DNA
in the spectral range of 300-800 nm2, d finally the collective symmetry
or genetic code whose nature appears to be artificial.3 Understanding and
explanation of the Kirlian photographs containing restored images of the
removed parts of plant leaves present even more difficulties.4'5 This
phenomenon is usually referred to as phantom leaf effect and can apparen-
tly be an indication of some unknown important features of genome.'6

It is quite obvious that the striking success of genetic engineering

can only emphasize the fact that the well known linear genetic code rep—
resents the protein synthesis code rather than the program of construct—
irig multidimensional eucaryotic organism. This disagreement between gene-
tics and embryology was predicted by A.A.Liubisohev7 and A.G.Gurwich8
long before the above mentioned phenomena were found out. The contradict-
ion became more significant when the problem of genetic degeneration of
some nations arose due to the enormous environmental pollution whereas
biological restoration techniques for radiation and chemically damaged

280 / SPIE Vol. 1621 Optical Memory a,xJ Neural Networks(199 1) 0-8194-0761-5/91 /$4.00

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 chromosomes are absent. This represents a strong motivation to develop
new approaches to morphogenesis that would give relatively consistent
treatment of the new experimental data and, which is more important, wo-
uld provide a clear mechanism or developing an embryo to an adult orga-
nism. This paper presents some experimental results on phantom leaf er-
feet (Section 2 ) and field—induced morphogenetic translations (Section
3 ) . The experimental data are treated us ing a new approach to morphogene-
sis which combines some physical models or holographic associative memory
and mathematical formalism o Fermi — Pasta — Ulam (PPU) recurrence ror
solitary waves in deoxyribonucleic acid (DNA).6
2 . ASSOCIATIVE MORY CONSIDERATION OF PHANTOM LEAF EFFECT
Phantom lear errect is usually observed in Kirlian photography which
is a type or electrophotography implemented using high voltage high £re-
quency pulses applied across the object to be photographed.4'5 In these
experiments a freshly plucked leaf is sandwiched between two metal elect—
rodes together with transparent insulation and a photographic film (the
detailed desoript ion of the experimental arang, regimes and results
can be round elsewhere.5'9) With a pulse voltage or specific waverorm ap—
plied to the electrodes a bunch o discharges channels out through the
lear, and an aura produced thereby is photographed directly on the film.
ir the leaves were damaged berore the exposure, in some cases (about 5%
or the trials) the developed photographs contained restored images of the
inner structure and the profile of the removed parts or the leaves (Fig.
1 ) . The phantom images can be also obtained with no discharge present
when only ultra weak endogenous radiation or the leaves is registered.

When analyzing the nature or phantom images we used the concept o!

c
chromosome field which controls the embryo morphogenesis and the adult

'1

a b
Fig.1 .Phantom erfect on a birch lear: a - photograph containing Kirlian
phantom image o the removed part o the lea! edge (rragrnent I ) ; b -
schematic or the sample preparation. No phantom is observed in the remov-
ed rraents 2 and 3.

SPIE Vol. 1621 Optical Memory arid Neural Networks(1991) / 281

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 organism structure.6 In the rramework of this approach the phantom gene—
ration can be considered as an attempt o the organism (leafl to reconst—
ruot and regenerate its lost parts, which demonstrates associative pro—
perties of the chromosome held.5'6 To explain this associative behavior
we use the well known holographic associative memory To
this end, we assume the morphoenetic inrormation to be stored in the
:rorm o holograms distributed in so called redundant or selfish DNA which
is estimated to be 95—98% o genome in higher biosystems.3 In ract, the
information capacity or the chromosome DNA molecule is actually high eno—
ugh to store not only 1—D genetic codes but the complete scenario oI the
organism development in space and time. We now discuss possible hologra-
phic mechanisms 01 morphogenetic information storage and retrieval with
attention to specific issues of coherent radiation sources, nonlinear
storage medium and and holographic memory configurations which can be im-
plemented in biological systems to satisfy the requirements of stability
and reliability or the restored fields.

2 .1 . Holographic model oI morphogenet Ic Inrormat ion storage and retrieval

To solve the problem of coherent radiation sources and, to some ex—

tent, of possible holographic recording and restoration configurations,
the concept or local rererences4 can be involved.6 This concept is based
upon the peculiarities or scattering electromagnetic waves emitted by a
single atom, e.g. a heavy metal embedded in a DNA molecule. Consider a
metallo—organic molecule with a single heavy metal atom in a well defined
orientation, which is typical ror liquid—crystal DNA. Irradiate the mole-
cule with X-rays slightly above the K-ecge o the heavy metal and observe
the angular distribution or the characteristic X—rays emitted by it. In
the absence o surrounding atoms, this radiation represents a spheric wa-
ye, which is assumed to be monochromatic. This assumption does not rest—
rict the validity o the consideration as can be seen below. In the pre—
sence or the surrounding atoms some o the emitted photons are scattered.
Assuming the presence or discrete atoms at positions r and weak scatter—
ing (see Fig.2b), the total electric field at the point or observation,
at a distance R, can be written as

E=AR + lff,(Of,)e t1,(1—cosO)

(1)

where I = 27tA = w/c is the wave number and A is the complex amplitude,
LL5 the scattering length (possibly complex), f,(O) is an angular dist—
ribution. We can make now an important observation: the angular distribu-
tion or the characteristic radiation contains a rringe pattern, which is
a hologram or the surroundings of the emitting atom. The rringes are o
macroscopic dimension being always in the ar field. This configuration
is usually rererred to as Fourier—Frauenhoffer hologram.15 The fringe vi-
sibility decreases rapidly with the distance or the scatterer rrom the
emitter. The rringe pattern has an angular width of about in

282 / SPIE Vol. 1621 Optical Memory and Neural Net works(1991)

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 the direction ol r, and in the perpendicular direction. In case
many similarly oriented molecules are present, each metal atom emitting a
photon produces a hologram of its environment. The waves emitted by die—
rerent atoms are incoherent and thererore do not interrere. Thus, ir the
metallo—organic molecules (crystallites) are oriented within an angle
(typically 1°) all the individual holograms are in register within .
Therefore their incoherent superposition produces a single hologram that
resolves the environment within r, << 2/3. Also, the crystal size, L,
should be smaller than R /r. Neither or these conditions imposes any
practical limitations on the method. Note that, compared to ordinary ho-
lography, the coherence requirement or the source is easy to satisry. The
periodicity of the crystal is immaterial, so the holographic experiment
could be done equally well on a liquid crystal (e .g. DNA) , a stretched
riber, or a well oriented membrane.

The approach or local rererences can be easily extended to randomly

assorted metal atoms ejecting photoelectrones and Auger—electrons. Some
of the ejected electrons get scattered by the atom's environment produc—

to Eq. (1 ) . I!
ing the electron angular distribution described by the equation similar
this electron distribution is recorded an electron holo—
gram or the emitting atom's closed environment is obtained. This mecha—
nism is or particular relevance to the phantom lear errect.

Let us discuss now the holographic recordin,g and reconstruction pro-

blem. Suppose the hologram is linearly recorded on a sphere or radius R.
Using Eq. (1 ) and assuming a strong reference wave or a weak scatterer, we
can easily get the hologram transmission which is equivalent to the La—
mous Gabor's hologram.15 Indeed, being irradiated with a spherical inco-
mine wave, which is the complex conjugate oL the reLerence, this hologram
reconstructs the Lield inside the now empty sphere: the light intensity
inside the sphere reproduces the reLerence beam and the scatterer. IL the
first term is filtered out, the intensity of the second term reproduces
the object, within the limitation oL the uncertainty principle. Note than
IL the hologram is recorded on the Lull sphere, there is no doubling oL
the image (only one real image is reconstructed) and the resolution oL
about is achieved in all three dimensions.

Unrortunately the ideal recording and reconstruction conditions are

not Lulfilled even Lor linear medium response and stable recording/retri-
eval conrigliration. Since the reLerence and. reconstructing waves acquire
an additional modulation due to the scattering on the object, a more ge-
neral holographic reconstruction problem is the Lollowing. Assume that a
compi bated , but known (reLerence ) waveLront interferes wi th an unknown
(ob ject ) wave , and the intensity oL the interLerence pat tern is recorded
on a sphere. Find an algorithm to reconstruct the source distribution of
the unknown wave, whose intensity can be written symbolically as
I = r(R + O)(R* + Q*) = r(RR* ÷ 00* + R0* + R*0) , (2)
where r,R and 0 are the reconstructing, reference and object waves res-
*
pectively, symbol denotes complex conjugate. The two principal terms,

SPIE Vol. 1621 Optical Memory arid Neural Networks (1991) / 283

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 rRO* and rR*O can be recognized as real and virtual images or the object
only 11 r = R* or r = R, respectively. Moreover, br both cases, the re—
constructed waveront in the rar field. represents the Fourier transrorm
YCI} or the field observed just behind the hologram containing the compo-
nent y{rRO* rR*O}, which is the convolution or the reconstructed object
with the autocorrelation o the rererence. Hence, undistorted image res—
toration is possible only Lor RR} 6(r) (delta—correlated rereren—
ce5 ) . And even in this case the other terms o Eq. (2 ) will introduce no—
ise, arrecting the accuracy or restoration. Thus Eq. (2) constitutes an
ill posed problem due to imperrect knowledge 01 the hologram impulse res-
ponse, i.e. the auto correlation RR}. It can be solved using optical
data restoration technique, e.g holographic associative memory systems
with feedback and iterative data recall.1°12 An outstanding merit or
such systems is their explicit and close analogy to Human brains, artiri—
cial neural networks°'" and the phantom leaI erIect.5'6

2 . 2 . Holographic assoc lat lye memory approach

An optical system with associative properties analogous to those or
phantom lear erreot is illustrated in Fig.2a. The memory contents is sto—
red in superimposed Fourier holograms, each or them being equivalent to
the holographic matched rilter or a stored image am recorded with angula-
ny multiplexed plane rererence waves Bm The memory also includes opti—
cal Ieedback and nonlinearities in the correlation domain to improve sys-
tem perroniance. To this end, the phase conjugate mirrors (POMs) '7 and 8
are positioned in the object and rererence legs oI the holographic rilter
optical system.

Suppose the system input is an image a, representing partial or di-

storted copy or one or the stored objects a,,. Then, a set or partially
reconstructed rererence beams B . is generated in the Fourier plane which
corresponds to b,, = *
am a
inthe correlation domain.10 Thus, each re—
constructed rererenee beam is weighted by the correlation or the input
object with the stored object related to that particular rererence beam.
The rererence beams are phase conjugated by POM 7 and rerlected back to-
ward the hologram, which reconstructs all the stored objects. In the re—
sonator configuration or rig.la, the reconstructed object waves are phase
conjugated by POM 8 and incident back to the hologram, i.e. the feedback
loop is closed. The process is iterated until a self—consistent solution
of the system is found. The complete set of the solutions or eigenfuncti—
ons of the system is merely the stored object set. Nonlinearities in one
or both of the POMs tend to form the basin of attraction in state space

([
around the stored objects. The output of the associative memory after POM
8 (first iteration) is given by
= ( * a) ® bm]) * ® am,), (3)

284 / SPIE Vol. 1621 Optical Memory arts-i Neural Networks(199 1)

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 where a are the stored objects, b are the amplitude or their associated
rererences in the input plane, K.eand "2 represent point nonlinearities
o! PCMs 7 and 8, * and ® denote correlation and convolution, respective—
ly. The double sum over the object subscripts m and. m' is due to the
light beam passing twice through the hologram. L the rererences are
assumed to be angularly multiplexed plane waves, the corresponding dist-
ributions in the input and correlation planes are spatially displaced
delta runotions, i.e. point light souTces. The separation As,Ay between
4 5
8

7P

a)
9

b) Fig.2. Similarity between optical and bi-

R ological holographic associative memory:
0
a - optical associative memory with mul-
tiple—iteration recall; 1 — incomplete
E input image; 2,3 - beam splitters; 4,6 -
I

lenses; 7,8 — phase conjugate mirrors; 9

— system output; 10 — rererence beams;
PHANTOM
b - phantom image rormation; R ,R are
DNA' :-
4E4
E I
local rererences; E,E2 are OPO pumps;
%%I
E , E are s ignal ( scat t ered ) waves ; E ,
L_..
E"4 are the conjugat e waves (phantom).

SPIE Vol. 1621 Optical Memory and Neural Networks(1991) / 285

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 rererence functions bm must be large enough to separate spatially dirre—
rent correlation/convolution terms in the rererence leg. L the nonlinea—
ri t ies K and 2 are properly chosen ( e .g . , thresholding ) , then the
selr—consistent solution a, will ultimately be reached arter several
iterations, which is the nearest to a in terms of correlation distance
measure.

In a simpliried single—pass single—POM configuration ot the associ—

ative memory,10'2 a dirruser is placed in contact with every stored ob—
ject a in order to sharpen the autocorrelation peak o the object rela—
tive to the sidelobes and to spread the image inrormation over the entire
recording area or the hologram. The dirruser encoding also results in
suppressing the crosscorrelation noise and in improving the SNR. Taking
into account that most biological objects possess the optical properties
or dirfuse scatterer, and that they can reproduce in fine details their
structure during growth and development, we hypothesize that the diuse
character or the light scattering may play an important role in the mor-
phogenesis. The intrinsic and individual dirruser encoding or the biolo-
gical structures can provide not only high inrorination capacity and reli-
ability or the morphogenetic data storage but may represent some kind or
key code which enables the system to recognize and adopt objects (i.e.
proteins) of its own structure and reject the objects or other structure.
This can be treated as a holographic concept oI immunity.

Since in the optical system or Fig.2 the hologram is imaged onto the
POM 7 in the correlation leg and object plane is imaged onto the POM 8,
the POMs can be placed directly in the hologram and object planes respec—
tively. In this conriguration, the DNA molecule represents also an appro-
priate medium ror rour—wave mixing, which is the main basis ror POM in
the considered model. In ract, the DNA possesses excimer and exciplex
states, which provide the population density inversion and laser errect
in DNA pumped In vtvo by metabolism or a cell.3 The pumped DNA constitute
a nonlinear medium and two conjugated pump plane waves are rormed in the
far field by any two neighboring local rererenoes between them (see rig.
2b). This conrigaration is similar to double optical phase conjugation
scheme in which two mutually incoherent counterpropagating pumps exist.6

Selr-sustained waves (autowaves) give another way or inrorrnation

transmission through the active media with the population density inver--
sion.i7 Two pump waves propagating through the active medium become seir—
oscillating, their waverorms being determined by the boundary conditions,
i.e. both by the hologram stored in DNA ("micro—structure") and, on the
other hand, by the object macro-structure. Such a system is highly redun-
dant, because the stored inrormation is distributed and multiply repli-
cated. Thus, any perturbation or the system parameters immediately produ-
ces an additional modulation or the rield, which is intended to compensa—
te ror the distortions introduced in the system structure. We assume the
phantom or the removed part or a lear to be an image or that very probe
and correction rield, which is also expected to occur in the rorm or ho-

286 / SPIE Vol. 1621 Optical Memory arid Neural Networks (1991)

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 lographic regulat or riel d during the embryo development.
In conolusion let us address some issues to be resolved in conriecti—
on with the justification of the holographic model or genome. First we
should point out some dirrioulties in describing and building the model
or a holographic associative memory in which both the recording medium
and the input object are uniformly distributed in some volume. The wave-
guide hologram appears to be most adequate to this conriguration.° This
analogy, however, calls ror a more detailed study and justification. Ano-
ther problem arises in considering mapping properties or holograms. A 2—D

phism ) o
or 3—D hologram usually interconnects two planes (input/output isomor—
the holographic memory opt ical system . Even in the case o a
2—D object this hologram (i.e. interconnection matrix) is described by a
4—tensor.°'" Since a volume hologram allows clean interconnection with
only three dimensions, the dimensionalities of the input and output ob-
jects cannot sum to greater than three, e.g. both input and output can be
3/2—D rractal objects." Note that the transmittance or an ideal difruser
is also a (3—ct)—dimensional 1ractal (0 < a < I )•i9 has been recently
pointed out that the DNA molecular structure displays £ractal dimensiona-
I
lity as well.20

3 a SOLITARY-WAVE MODEL OF MORPHOGETIC TRANSLATIONS

As has been already mentioned, the codirig o the biosystem structure
may proceed in the rorm o autowaves and solitary waves, e.g. , breathers,
whose internal oscillatory structure is associated with a hologram in
space and time domains, which stores and reconstructs specific spatial
and temporal status oI a developing or regenerating organism.6 The rorma-
lism o this version is based upon the Permi-Pasta—Ulam (FPU) recurrence
phenomenon, which represents time periodic return o the energy spectrum
or the initial perturbation in a distributed non—linear oscillatory sys-
tern into its primary state without thermalisation.2 It has been shown
that the DNA molecule can be treated as a FPU resonator.6 In this model
the dynamics o electron density wave, distributed along the sugar—
phosphate chains of the DNA is described by the non-linear ScIu'odinger
equation. Hence, the electron density oscillations in the nucleotide
structures can be considered as excitation point sources, uni!orrnly dist—
ributed along the sugar-phosphate chain, representing a kind or a long
electric line.

3.1 . Experimental morphogenetic translations

The proposed model of organizing genetic material can be userul ror
understanding the nature 01 biosystems distant interaction. Experimental
investigation of distant interaction of biosystems was carried out using
the primary embryo induction technique. Any induction includes two compo-
nents : inductor t issue (donor) and induced part , . e • , target t issue (re-
C ipient ) . The donor t issue should be able to arrec t in some specific way
i
the recipient tissue and stimulate its development. In turn, the recipi-
ent should possess surricient competence to respond to this action and

SPIE Vol. 1621 Opt/ca/Memory ariNeixalNetworks(199l) / 287

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 develop through corresponding embryo stages.
It was previously assumed that informative interactions between emb-
170 tissues are possible only in mechanical way, e.g. due to the tension
o:r cell membranes. A number or simple experiments show, however that emb-
ryo induction can ooour also without mechanical contact oI the tissues.
Extraots rrom certain biological tissues acting on the recipient results
in morphogenes is or corresponding embryo s true tures . Thus , morphogenes is
(cell dirrerentiation and struoturogenesis) can be caused by the chemical
action or specific morphogenetic substances whose influence on embryo is
not clear either. Our model or morphogenesis allows us to assume donor
and recipient tissues to interact by means or specific field generated by
donor and modulated by genetic information in symbolic form. These fields
can be also produced by morphogenetic substances.

In the experiments, the broad—band electromagnetic field was gene-

rated by specially designed electronic FPU—oscillator with a spatially
distributed nonlinear resonator. It simulates hypothetical runotion o
eucaryote chromosomes namely readout and translation of morphogenetic in-
rormation £rom a certain biosystem to the genome of another biosystem be—
:Lng taxonomically identical to the first system. to veriry this model we
used the donor biosystem representing a tadpole of Xenopvs Iaevts (a
frog) at the stage NP 44—46 and the recipient biosystem being an ectoderm
or early gastrula (outer layer of an embryo at an early stage of develop-
ment ) or the same rrog at the stage NT' I 0 (the stages or embryogenesis
have been classified accordirig to23). Microsurgery operations, tissue cu-
ltivation and morphologioal analysis have been performed using conventio-
nal technique. Morphogenetic information translation was realized with 4
donors present inside the FPU-oscillator cavity and 24 accepters placed
in Petri dishes with culture medium which were situated at distances 25
cm to 2 meters rrom the oscillator. The translations occurred when the
PPU-oscil lator was switched on during 5 minutes . The control experiments
were made by switching on the oscillator for 5 minutes with no donors
present.
The experiments demonstrated the possibility of morphogenetic inror-
rnation transfer by the broad-band electromagnetic field modulated by the
living tissues o tadpoles or Xenopua laevts (donor) on the cell dirfe-
rentiation o an embryo tissue or the same species. In some cases (about
1%) the embryo tissues developed into the structures containir&g the comp—
lete set o mesoderm and neural derivatives or the primary embryo tissue
(Fig.3). The dirrerentiation was not observed in 100% of control experi—
ments in which the donors were subjected to the action or "pure" broad—
band electromagnetic field or the FPU-generator with no inductor tissue
present. The results obtained confirm the soliton—holographic hypothesis
of eucaryote genetic mechanism and camiot be explained in the framework
of tradi t ional concepts of biological morphogenes is.

3.3. DynamIc Light Scattering In DNA

Nonlinear dynamics of chromosomes including self—oscillations and
solitary waves in DNA are relevant to the proposed model of morphogene-
sis. Thus, the search for actual nonlinear phenomena in DNA was an aim of

288 / SPIE Vol. 162 1 Optical Memory arid Neural Networks(199 1)

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 a b
Fig.3. Photographs or a nerve fiber (a) and muscle tissue (b) which were
developed. rrom ectoderm o an early gastrula o xenopu Zaevts. Magnif 1-
cation is 10x40.

this work. In studying nonlinear properties o! informative biopolymers we

used the photon correlation spectroscopy technique. For this purpose the
Malvern spectrometers were applied along with correlators 4300 and K7032.
The experimental conditions were similar to those described elsewhere.2'
The samples or native highly polyrnerized DNA (restricted rragments up to
100 base pairs) rrom calf's thymus were used. One-strand DJA was obtained
by conventional 10 minute heating in a glass vessel at 100 C in a water
bath with subsequent rapid cooling in liquid nitrogen. The control expe-
riments to check possible background vibrations o the cuvette were with
a simple o silica gel possessing a strong light scattering.
In the experiments, a variety o quasi—periodically repeated selr—
oscillating correlation runotions were obtained, which demonstrated the
S imilarity to FPU recurrence (Fig .4 ) . Most excit irig are phantom autocor-
relations observed in some cases after removal or the DNA samples £rom
the cuvette space (see Fig.4c). Unfortunately, we cannot reasonably acco—
unt ror this phenomenon, which, however, appears to mimic the phantom le-
ar eirect.

4 S CONCLUSION

In this paper, we have described several experiments, which can be

explained in the framework o the proposed soliton/holographic approach
to morphogenesis. The underlying principles o! ho1ographic storage and
solitary wave transfer of morphogenetic iMoration reveal some new as-
pects of genetic runctions in biological systems which have never been
known berore. This new insight into the nature or morphogenesis makes it
possible to treat genome as a biological-holographic computer which gene—
rates endogenous solitary electromagnetic and acoustic waves to carry 4-D
epigenetic inrormation used by biosystems ror their spatial and temporal
selr-organizing.

SPIE Vol. 162 1 Optical Memory arid Neural Networks(1991) / 289

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 VLUERN (<( )C732 ))) Ve1bson 2.1 Date -1--199 Ti 16:18:6
Conielatorl Fundaisental Saisple Tisse (j) =
Auto-corie1 ati
2.3E+4i

1.?E434L
: —
f.5 T::
;i1
Seia1 Configuration Channel tusbei'.
Title:-TEST DATh C

The angle o scattering is 60 , t

Fig.4. Autocorrelation runetions o the light scattered on a DNA sample
(hard gel ) in the cylindrical0 cuvette oI 1 cm diameter and 5 cm height.
= 2000 s per channel (correlator
4300). The repeated runctions of plots a and b are obtained at 6 and 22
minutes arter the beginning or the experiment. Intermediate runctions
dramatically dirfer from the functions or plots a and b. Plot or Fig. C)
contains the phantom correlation obtained in spectrometer "Malvern—K7032"
after the cuvette with DNA sample was removed rrom the cuvette space.

5 . ACKNOWLEDGME2ITS

The authors graterully acknowlecige a very heiprul assistance or A.

G. Kobzunenko, 0. A. Khaperskaya, M. E. Bogdanov and 0. V. Stolbovskaya
in the experiments.

6 . REFERflWES

I . B . B . Khes in , Genome mob t 1 t tj , Nai,&a Pbs . , Moscow , I 984 ( 1 Ti Ru88 I art).

2. V. I. Scherbak, "Collective Symmetry or Genetic Code," Jotirri. of The—
or. Biology, vol. 132, pp. 121-124 Jan. 1988.
3. F. A. Popp, "Coherent Photon Storage or Biological Systems," In EZec—
tromagnettc B1.o-trtformattort, F. A. Popp, U. Warnke, H. L. Konig, W.
Peschka , Eds . , Urban & Schwarzenberg, Pbs . , Munohen, 1 990 pp . 1 44-1 67.
4 . J . K . Chouclhury , P . C . Ke jariwal , and A . Chat topadhyay , "Some Novel
Aspec ts or Phantom Lear Errect in Kirl ian Photography , " Journ . of the
Inst Itutton of engineers, vol. 60-EL3, pp.67-73, December 1979.

290 / SPIE Vol. 1621 Optical Memory arid Neural Net works(199 1)

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms

 5. 2. P. Gariaev and A. M. Yunin, "Pact or Phantom? (Again about dying
leaves mystery)," Energij .* 10, pp.46—52, October 1989, (In Russtan3.
6. V. P. Kaznaoheyev, P. P. Gariaev, A. A. Vasiliev, and. A. A. Berezin,"
Soliton-Holographic Genome with Collective Symmetry o Genetic Code,"
In8tttute of CZIntcal and Experimental Medtcine, Siberia Branch or
the USSR Acad. or Sciences , Novosibirsk , June 1 990 , ( In Russ tan).
7. A. A. Liubischev, About the Nature of Herecittj Factors, Gosizdat,
Perm, 1925.
8. A. G. Gurwich, Theory of btologlcal field, Gosizdat Pbs. , Moscow,
1944.
9. J. K. Ohoudhury, P. C. Kejariwal, and A. Chattopaclhyay, "A High Vol-
tage High Frequency Pulse Generator ror Kirlian Photography, " Journ.
of the Inatttutton of Engineers, vol. 60-EL3, pp.67-73, December
1979.
10. Y. Oweohko, G. J. Dunning, E. Marom, and 13. H. Sorrer, "Holographic
associative memory with nonlinearities in the correlation domain,"
Appi. Opt. vol. 26, pp. 1900-1910, May 1987.
I 1 . H. J. Caulfield, "Optical Neural Networks , " Proc . IEEE, vol.77,
pp.1573—1583, October 1989.
I 2 . A . A . Vas il iev , E . I . Galavanova , I
. N . Kompanets , and G . A . Luniako-
va, "Holographic Associative Memory With Liquid Crystal Light Valve
as a Phase Conjugate Mirror , " Lebedev Fhjs . Inst t tute Prepr tnt , I 84,
FIAN, Moscow, 1990, (In RU281afl).
1 3 . Evolut tori of genome, Mir Pbs. , Moscow, I 986, (tn Rth9sLan).
1 4 . A . Szöke , "X-Ray and Elec tron Holography Us ing a Rererence Beam , " In
Proceec1lng8 , Short Wave-length Coherent Radlat I on and AmpZ tfI cat Ion,
pp. 361—367, 1986.
15. R. J. Collier, C. B. Burckhardt, and L. H. Lin, Optical Holography
Academic Press , New York , London , I 971).
16. S. Weiss, S. Sternklar, and B. Fisher, "Double Phase-Conjugate Mir-
ror: analysis , Demonstrat ion and Appl icat ions , " Opt . Let ters , vol.
12, pp. 114—116 Jan. 1987.
17. G. R. Ivanitsky, V. I. Krinsky, and O.A.Mornev, "Autowaves: an Inter-
disoiplinary Finding, " In Cybernet Ics of 1 lying matter: nature, man,
I nformat I on , I
. M . Makarov Ed . , Mir Pbs . , Moscow , 1 987 , pp . 52-74.
18. D. I. Mirovitsky, I. P. Budagian, and V. P. Dubrovin, MtcrO-tLX2veLIde
opt Ics and hoIographj, Nauka Pbs . , Moscow, I 983 (In Ru8sIan).
19. Ya. B. Zel'dovioh and D.D.Sokolov, "Fractals, similarity, an interme-
diate asymptotics," U8peP.hI Phjs. NauJ, vol. 146, pp. 493-506, July
1985 (In Russian).
20. N. A. Bul'enkov, "Seir-organizing Triplet Structures or Ideal Frac-
tals in "Bound" Water with the Symmetry ID3 and T," Crj8taIograph,
vol.35, pp. 147-154, January 1990 (In Russian).
21 . E. Fermi, J. Pasta, and S.Ulam, "Studies or Nonlinear Problems," In
Col lected Papers of Enr I Co Fermi , Vol . 2 , University or Chicago Press,
Chicago, 1955, pp. 978—988.
22. P. Gariaev, A. Berezin, V. Ohudin, and M. Yalakas, "Chromosome Bio—
computer, " Vrach, .* 4, pp. 30—33 , April 1 991 , ( in Russian).
23 . Objects of the deveZoprnent biologij, Nauka Pbs . , Moscow, I 974 ( in Rus-
sian).
24. S. S. Sorlie and R. Pecora, "A Dynamic Light Scattering Study or Four
DNA Restriction Fragments," MacromoLecules, vol. 23, pp. 487-479, Fe-
bruary 1990.

SPIE Vol 1621 Optical Memory a,xi Neural Networks(1991) / 291

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 12/03/2013 Terms of Use: http://spiedl.org/terms