Current human over-population and the logically consequent expansion of novel viral diseases combined with the widening spread of antibiotic resistance among the “traditional” human pathogens calls more and more for inexpensive, rapid and universally-applicable detection methods – capable of detecting wide-variety of pathogens in a single test run. The epidemiological consequences resulting from the increased world-wide mobility of the human population is also demanding for disease detection novelty – which is well demonstrated by the rapid progression of viral infectious diseases eclipsing with the social and economic amok caused by the COVIR-19 disease.

The current state of medical diagnostics is incapable to serve this purpose because it relays upon traditional technologies employing (1) antibody tests (that are both too specific to single-type microorganisms yet, either extremely specific or too low-specific to the detection of sub-varieties of particular species; they simply lack the high detection precision of DNA-based detection) or (2) Polymerase Chain Reaction (PCR) based tests that are highly precise and can be moderately flexible to cross-species recognition, yet they require both time- and equipment investment to be able to become comfortably field-employed.

Although the above assessments seams to hint at a dead-end road, there is amazing technology that has lately become forgotten regardless the fact that it was developed more than 30 years ago. And the reason for that is, because its capabilities failed to be well-argued and widely proven, and because very-few scientists were experts in the field with the rest of the community being under-proficient to the scope of its applicability.

This technology was called “Sequencing by Hybridisation” (SBH) with its off-spring - DNA-chip technology. Out of many serious attempts during the 1990s to commercially develop and justify its economic sense, practically only very few companies survived to still exist in the field – with questionable profitability and murky future. Although development started over 25 years ago, efficient practical implementation of the microarray technology is lagging behind its advances and immense medical and commercial potential. Current expression arrays are still expensive and inflexible. They are custom-designed for each organism and they do not offer the possibility of incorporating updated genomic information without production of a new chip. Note, however, that a patent has been issued allowing to use all data generated by any-kind of DNA arrays and re-analyzed in different prospect and means to allow for dramatic increase in scope and usefulness of the information incorporated in those chip-based data results (U.S. Patent 7,031,745); that is at least 10-fol increase in informational data.

Nevertheless in 1993 there was one research “off-spring” that was apprised by and granted the intellectual rights from US DOE and Argonne National Laboratory (ANL) [DOE Case No. S-82,524, ANL-IN-94-109 “DNA Library Characterization By Oligonucleotide-Based Clone Fingerprinting (Coding) that was able to prove for the first time in the world the mind-blowing ability of the SBH and Chip-based technology approach developed by Dr. Chris Dyanov for practically unlimited detection capability coupled with unlimited inter-species detection ability – so much so, that, if precisely designed, it allows for the detection of the existence of unknown microbial pathogens regardless of the species host they may reside in. Today, all this sounds so much too-good-to-be-truth – as much as it did back in the 90s. The reason the technology remains undeveloped for over 30 years is that it sounds pretty-much as a science-fiction and because its unusual complicity makes it difficult to be communicated and to become correctly understood. Yet, we have all scientific data available to provide for the understanding of its true capabilities being practically tested and verified long time ago. We, and some others had called this technology Gene Fingerprinting by Hybridisation (GFH).

The Gene Fingerprinting by Hybridisation (GFH) is an off-spring of the SBH approach first developed in late 1980s by the mathematician William Barns [Bains, W. and Smith, G.C. A Novel Method for Nucleic Acid Sequence Determination. J. Theor. Biol. (1988) 135, 303-307] and tested experimentally in 4-6 laboratories in the World at that time. The version of SBH we have pioneered and employed is quite simple as an idea, but requires significant both research-and-development and computation resources, both of which are not limiting factors today.

Simply, the idea is this: If one design, chemically synthesize and manufacture on a chip a relatively modest number of proper DNA molecules (say, several thousand up to a 100,000) – that chip will be capable to provide for creating a fingerprint of extremely large amount of genes regardless of which- and of how many different species the gene originated from – that being achieved by both bio-chemical and/or electrochemical detection, and computational analysis. Thereby, this technology can serve as unlimited and highly-precise detection approach serving both medical- and National bio-security detection applications. We were able to prove practically the capability of this technology as yearly as during the period 1993-2003.

Below we will describe and present some results obtained by the application of this technology without disclosing too much precise proprietary information and with some simplifying in order to make it easier to understand.

But first, a short introduction will be given in order to elaborate an idea as to the scope and the challenges faced at the start of this project.

Every feature in every organism in Earth from the simplest ones (viruses) to the genetically most-complicated ones (plants) is encoded, expressed and passed in generation by chemical bio-macromolecules called RNA and DNA (ribo- and deoxyribo-nucleic acids). Within their chemical structure, four molecular features, called “bases”, serve as the code letters (named A,U,C,G in RNA and A=adenine),T= thymidine, C=cytosine ,G=guanidine in DNA) that are capable of recognizing each-other in pairs of complementarity A-T (respectively, A-U for RNA) and C-G, as well as being recognized biochemically by the cellular machinery to proceed with implementing the base-encoded information into life biochemical events and structures.

Figure 1: Structure of a DNA molecule. (from: Principles of Cell Biology (BIOL2060). Available online: http://BIOL2060/BIOL2060-18/18_04.jpg or the link here.

Smallest-size genome identified is from a Viroid family (viroids are the smallest known pathogenic agents) and one of the smallest belongs to the Grapevine yellow speckle viroid with 220 nucleotides only, and the Rice yellow mottle virus satellite with 220 bp (base-pairs); then Avocado sunblotch viroid (ASBV) – a linear genomic RNA with 247 BP size (https://www.ebi.ac.uk/genomes/viroid.html).

Let’s also mention some of the smallest organisms we know – bacteria. The smallest bacterium yet sequenced comes in at only 112,000 letters, and was discovered living inside of a tiny insect. A common and well-studied bacterium, E. coli, has a genome of around 4,600,000 nucleotides - only about 1/1,000th the length of the human genome.

With only 4 different letters, this information this seems simple enough – until you realize that the human genome is about 3 billion letters long! That’s a lot of possible ways to arrange those 4 nucleotides. With the exception of identical twins, every human being has a slightly different order of letters in their genome. And just one single cell - a skin cell, let’s say, – contains one copy of that entire genome. 3,000,000,000 is a lot of letters. If one considers that a typical 250-page book has about 500,000 letters, that means - it would take 6,000 books to hold one person’s genome. But the human genome is nowhere close to the size of the biggest genome out there.

The record is currently held by a rare Japanese flower named Paris japonica, coming in at 150 billion nucleotides – 50 times the size of the human genome! Next biggest animal genome belonged to the marbled lungfish with 130 billion base pairs.

Figure 2. Paris japonica, the rare Japanese flower that holds the current record for largest eukaryotic genome at 149 billion nucleotides. Eucaryotes are species in the cells of which the nucleus including the DNA is separated from the rest of the cell content by a membrane; these are more evolutionary-advanced species than the most microbes (i.e. the prokaryotes).

How is it possible that a humble little flower can have so much DNA in one of its cells? Simply, because there is more to a genome than just the total number of nucleotide letters. Not all of those letters actually give instructions to make something. In fact, only a fraction of a genome actually has the information for building the chemicals the cell needs. One “coding” region, known as a gene, holds the instructions for how to make one protein. Even though the human genome is 3 billion letters long, it only holds about 20,000 genes that are known to be structural genes, i.e. to code for the synthesis of a bio-chemical molecule (a structure).

So, in the case of the Japanese flower with the massive genome, much of that DNA contains actually duplicate copies of the plant’s genes. In summary, more DNA does not necessarily mean more information - in this case, it is sort of like having 100 copies of the same novel rather than having 100 different novels. These copies are not necessarily meaningless, however - gene duplication can lead to novelties and effects that scientists are still working to understand.

In the animal kingdom, the relationship between genome size and evolutiona-ry status is not clear. One of the largest genomes belongs to a very small creature, Amoeba dubia. This protozoan genome has 670 billion units of DNA, or base pairs. The genome of a cousin, Amoeba proteus, has a mere 290 billion base pairs, making it 100 times larger than the human genome.

Amoeba dubia: 670,000,000,000 bp genome
Amoeba proteus: 290,000,000,000 bp genome

Figure 3. Microscope photo of Amoeba dubia (top) and Amoeba proteus (bottom) – the champions in genome size.

Whether it is a tiny bacterium inside an insect or a towering spruce tree, every living thing we know of - shares the same 4-letter instruction code. Humans fall in the middle of the pack in terms of genome size, but there is more to a genome than just the overall length. The part that actually gives unique instructions does not always match up with the total size.

While it is indeed a big deal, sequencing an entire genome of a living being is still just only a first step toward decoding the principles of genetic information storage and interpretation. Figuring out what all those letters really mean in the way they are arranged is an even bigger task, and many discoveries have yet to be made inside the genomes of Earth’s organisms.

Immense practical needs in the medicine and genetics (in diagnostics) will be served, if a practically simple and rapid technology is developed to allow for sequencing, or for code-based identification of genes obtained from live organisms.

Here, in simple terms we will explain the goal to be achieved and complications that needs to be solved.

1. Let’s imagine that one gene (encoded with only 4 letters) is a page-long sentence (2,000 letters long each, for a median size gene) and it belongs to a book with multiple “chapters” (chromosomes; 46 in total in a Human genome in 23 almost-identical pairs), and the book is 1.5 million (1,500,000) pages.

2. For the purpose of any practical gene analysis, all 1.5 million pages has to be shredded horizontally to separate rows of text with no spaces and punctuation between the 4 letters of code. That would result to a random rows’ mix of approximately 6 million 6,000,000 text-strips (if each strip contains 500 letters, or ¼ of a gene). Unfortunately, in the real practice, at least 1,000 cells are needed to obtain these fragments of DNA/RNA for a study, so in our basket to analyze, we now have not 6 million, but 6 billion (6,000,000,000) pieces of genetic fragments or “paper-strips”!

Figure 4. Simplified scheme of genome analysis principles – DNA probe preparation and oligonucleotide “fishing”.

3. The minimal goal is to achieve capability to detect and identify the presence of 1 page out of several millions based upon reading the information from single-row strips that belong from that “page” we here addressed as a gene. The technology for the minimal goal already exists, but it takes from few hours to few days and serious machinery to achieve a particular gene identification.

The maximum goal would be to identify each one of the several billion fragments that belong to the book (one genome) and, eventually, to restore all page numbers and sort each strip with letters into specific pales of strips – corresponding to the numbered page (gene) they belong to. Although there is some progress in this direction, no credible technology is capable of achieving this in rapid fashion, nor to a practically meaningful commercial extent.

a) We start with the presumption, that we have a humongous pile of strips we know nothing about. So, we have to sort as many pieces (strips) as possible into piles that belong to the same original page (gene) to the highest degree of confidence possible.

b) The only way to do that is based upon identifying (reading) fragments of texts we will name as “words” (in fact, oligonucleotides = small DNA molecules) containing a defined number of letters in each “word” as we choose to synthesize them so.

c) Here comes mathematics. Because we have a DNA-code of only 4 letters, if we choose the size of each “word” (oligo), we will “fish” to identify, to consist of only 8 letters (for example) of A, T, C and G, then all possible words there that can be written and synthesized is 4⁸=65,536 words. So, we will need to synthesize and arrange on a chip only 65,536 small molecules of DNA (called oligonucleotides, or “oligos” for short) in order to be used as a “fishing-hooks” to catch from the pile the exact strip, in which is the corresponding word existing. And that is a relatively small total number of DNA oligos, that could allow for an easy goal achievement. If we choose a word-size of 10 letters, then all possible letter combinations are 4¹⁰=1,024,000 words (i.e., oligos needed) as fishing hooks capable to allow for reading anything written on all and any page of the book - that becomes almost impossible to achieve technically and with precision.

The problem with using “fishing-hook” oligos (words) with only 8 letters (that is extremely appealing) is, that, first, it is almost impossible to maintain biochemically the necessary conditions to allow for “molecular fishing” and, second, it is not going to serve for a very-precise page identification and sorting computationally, because they will catch the same letters but belonging to different pages, i.e. they are too abundant (being only 8-letters in size) and the finger-prints of pages is not going to be very accurate; at least – not universally precise and accurate.

d) There is the main problem to be solved: we somehow have to design oligos (“fishing-hooks”) more or less longer than 8-letters each, but as short as possible and/or use a mix of different-size oligos but chosen with a knowledge of their abundancy among all genomes of all organisms. Thereby a pools of detection oligos should be informationally chosen after computati-onal analyses of sequenced genome databases to provide for both – highly unique oligos capable of specifically identifying particular genes (and gene- and function-features), and highly abundant oligos - capable of serving for points for gene fingerprinting analyses.

e) As a final result, the fingerprinting chip should be capable to “map” for a pin-point gene-fragment identification AND for fingerprinting of gene sequences to allow for gene-(sequence)variations mapping. The last feature would serve for probing the presence of yet-unknown genes of different level of similarities to known gene sequences. Implementing these approaches resulted in a patent co-authored by one of us (U.S. Patent 7,031,845)

In order to find the correct relative placement between all the DNA-fragments obtained from a (organism), some information on DNA clones (i.e. clones of the genes studied) overlaps must become available during the analysis. In the technique of oligos fingerprinting, short labeled DNA-sequences, or oligo probes, do attach via the phenomenon of complementarity pairing (or, as professionally called – they hybridize), to corresponding positions along the target DNA of the clone (i.e. gene-copy) matching their own DNA sequence (precise “fishing” - that is). For this purpose, the oligonucleotide probes are usually non-unique, i.e., they occur at many points along the genome (within different genes), and typically they will hybridize with 1-50% of all the gene clones in a studied sample. The figure below illustrates the hybridization scenarios (in red color) as being computationally aligned along a particular target DNA sequence:

Fig.6: Oligonucleotide-based Genome Investigation Strategies. In red are shown the specific places within the genome DNA fragments where labelled probe-oligos had attached during the hybridization molecular binding. Oligonucleotide-based fingerprinting strategy is using oligos to create a “fingerprint” for the identification alone of majority of gene fragments existing in a biological probe. Mapping is an approach not only to identify the genes in a probe, but to allow for their localization (mapping) onto a known (or unknown) genome map by exploiting the overlapping sequences of the used oligos being computationally reconstructed (overlapped); it allows for the creation of genome maps with relatively precise positioning of the genes within a genome (i.e. where are positioned the pages within a genome book). Sequencing by oligonucleotide hybridization (SBH) allows not only for fingerprinting and mapping the genes in a sample DNA mix, but for reading and computational reconstruction of the entire precise-identification of each letter and its position within a gene, and within the entire genome (i.e. precise reading\decoding of all gene information). For the purpose of SBH, all used oligo sequences must allow for their continuous and uninterrupted overlap exploited in computational reconstruction of the entire gene (genome) sequence.

Physical mapping is the process of determining the relative position of genetic markers (sequences) along the DNA. The resulting genetic maps are used further as a basis for DNA sequencing, and for the isolation and detailed characterization of individual genes or other DNA regions of interest. The construction of a high-resolution physical map of the entire human genome, and that of other organisms, is one of the top priorities of the Genome Projects. The mapping process involves the production of genetic clones, i.e., cloned pieces of DNA or RNA, each representing a chromosomal (genome) segment. The mapping itself is the reconstruction of the order of and physical distance between the DNA-clones – based upon the overlapping on the oligo sequences used for their identification by the means of the multipoint hybridization results. Below is an example of a physical map reconstructed by computationally aligning the position of each and all DNA-clone mapped onto the known genome or onto a newly-constructed genome map.

The short horizontal lines are the DNA clones with their x-coordinates corresponding to their position on the (original) target genome. The y-coordinates correspond to the DNA clone overlapping-order in the computationally constructed map. Note that each point on the target genome is covered by many clones – as is the real-life clone representation within gene libraries constructed by scientists. The total length of the DNA clones divided by the length of the genome is called the clone coverage (10 in this example on the figure). The location of the DNA clones along the target genome is not directly known to the experimenters – experimental mapping data, such as the oligonucleotide-hybridization data, is generally used in attempt to reconstruct the entire genome map. The solution to the mapping task is generating a sequence list for every DNA-clone that gives its estimated position on the genome map. A list of DNA-clones covering a continuous section of the genome together with their physical distances and sequences is called a contig. With a sufficient (synthesized and hybridized) oligonucleotide coverage, the whole genome map is usually one contig. Such a plot of DNA clone order in the constructed map vs. the real clone position provides a visual map-quality measure. Small errors, which do not change the clone order, cannot be seen from the plot. A completely random solution will correspond on the map to randomly placed (-aligned) clones, whereas a non-random solution containing several large errors will look like randomly placed broken contigs with approximately correct intra-contig order (see the Figure 7).

The methodological approach for oligo fingerprinting hybridization consists of the following steps:

1) Prepare oligonucleotide chip by chemically synthesizing short oligo probes and chemically attach them to a solid surface to allow to manipulate all of them simultaneously. The surface can be optically transparent or not, but to allow for optical acquisition of the hybridization result; or to allow for electrochemical acquisition – as a computer microchip. It could be made from plastic, glass, silicone or metal-coated surface.

2) Obtain and process DNA or RNA from a biological source in order to prepare solubilized pool of mixed DNA-fragments for their analysis by hybridizing them on the chip where each of them will bind to any and all oligos deposed on the chip if a corresponding sequence to an oligo of the chip is presented in perfect-match on a probed DNA-fragment (clone) in the solution. Of course, the biochemical conditions for the perfect-match binding (“fishing”) are mastered proprietarily a priori.

3) Design and explore conditions and measurements for precise quantitative measurement of each and all binding (hybridization) events on the chip in order to acquire complete data for all hybridization events to be used later in computational analyses.

4) Computationally manipulate all acquired hybridization data so to eliminate interferences and to achieve comparable and interpretable results for further computational analyses among multiple hybridization events, multiple different chips and multiple biological samples. For that purpose, different signal-scaling approaches and strategies are employed for data manipulation.

One of the crucial components of the oligo fingerprinting analysis is the process by which the signal intensity data are transformed into hybridization fingerprints. The hybridization fingerprints specify whether the oligo probes hybridize, or do not hybridize, to the DNA clones from the sample, and to what multiplication extent they do hybridize to a same oligo probe (i.e. to account with precision for multiple hybridization events in the chip). Oligo-probes that hybridize to the DNA clones should produce larger signal intensity values than those that do not hybridize in a perfect-match fashion.

The following is a description of how prior oligo fingerprinting studies have processed the data to produce the hybridization fingerprints. Hybridization fingerprints results have been generated by transforming the hybridization signal intensity data into three discrete values 0, 1 and N, where 0 and 1, respectively, specify negative and positive hybridization events and N designates an uncertain assignment. The signal intensity data from the unidentified DNA clones were transformed into 0, 1 and N based on the signal intensities from control clones, which are clones with defined nucleotide sequences and from measurements and scaling transformations among all measured hybridization events on the entire surface of the chip. For most oligo probes, the control clones expected not to hybridize with the probe (negative controls) have signal intensity values less than the control clones expected to hybridize with the probe (positive controls); conversely, the signal intensity values from the positive clones are higher than those from the negative clones. For probes that function in this manner, clones with intensity values ≤X were given a 0 classification, where X is the highest intensity value generated by a negative control. Clones with intensity values ≥Y were given a 1 classification, where Y is the lowest value generated by a positive control. All other clones were given an N classification. For some probes, not all of the control clones perform in the predicted manner; for example, some positive control clones may have intensity values that are lower than some of the negative control values and vice versa. For probes that function in this manner, clones with intensity values <X were given a 0 classification, where X is the lowest intensity value generated from a positive control. Clones with intensity values >Y were given a 1 classification, where Y is the highest value generated by a negative control. All other clones are given an N classification. Performing this analysis with all probes for all clones creates a hybridization fingerprint for each clone. An example of a hybridization fingerprint created by 26 probes is 000101N001000N110101111000. This is actually a bit simplistic explanation, because the real data processing calculation is by far more complicated involving scaling, rank-scaling and data reconstruction to account for various signal-acquisition- and experimental variabilities.

Computationally, hybridization-signal matrixes are constructed where the DNA-clones are the horizontal lines of genetic sequences (encoded by the nucleotides represented by letters A, T, C and G). The random occurrences of a single non-unique oligonucleotide (oligo) probe, are marked by the dotted vertical lines. If the oligo probe as an example occurs 3 times along this 7 clones genome section (1), its occurrence vector is digitally represented as (1,1,0,0,1,2,0). Under experimental conditions eliminating the hybridization noise (miss-match discriminating condition), oligo probe occurrences determine positive oligo-probe hybridizations. In such a case the hybridization vector of the probe is going to be (1,1,0,0,1,1,0). The hybridization finger-print of a clone is the row in the hybridization matrix corresponding to the hybridizations of all oligo-probes with the particular clone – combining their (oligo) sequences and hybridization signal values. Overlapping clones are likely to have similar fingerprints, and this fact is the basis for the recovery of the clones’ relative positioning among the genome, i.e. genome mapping (by reconstructing genome contigs).

In realistic scenarios, the level of experimental noise can result in both false-positive hybridizations (hybridization-detection with no oligo-probe occurrence), and in false-negative hybridizations (oligo-probe occurrence with no hybridization detected). Nevertheless, Bayesian statistics can still be used to identify overlapping clones, provided a sufficient number of probes is used. Moreover, certain computational and experimental techniques can produce significant level of detection accuracy and false-signal discrimination, for example by applying scaling and rank-scaling measurements of signal-to-noise rations.

Our methodology approaches and procedures were precisely optimized and tested, and a small-scale experiment was performed with 20 chip replicas each containing 27,500 cDNA clone fragments from human infant brain cDNA library. A ½ of a real hybridization result is shown below with signal-data visualization from a small row-sector as an example how data are accumulated before computational processing (Figure 8).

Figure 8. DNA chip images on a high-density nylon membrane array (31104 dots; 1/2-part of the image presented) obtained after hybridization with oligonucleotide probe (GCTCATCGTC) and a corresponding signal diagram, presenting the signal-intensity vs. clone-position relation - obtained from a measurement of a small part of only one row from each of both images from the same chip hybridized. Image A shows a membrane image after 4 hours exposure at 4°C; image B - after 14 h exposure; the longer detection exposure – the less signal noise interference is detected. All numbers indicated show some of the areas with highly positive cDNAs hybridized with the target oligo shown. K indicates 1000 cpm. The grid and small points were created after a computational measurement and optimization of the image by “DOTS” program (created by Jonathan Jarvis) designed for better signal alignment, dots'-position visualization and 2-dimentional separation.

Physical mapping using relatively non-unique oligo probes shares the advantages with the popular mapping techniques – that the use of unique probes offers much simpler experimental handling requirements, as well as the ability to analyze many DNA clones in parallel. It also enjoys the discrimination accuracy of DNA repeat-sequences oligos compared to the standard gel-based fingerprinting (sequencing). Its advantages over mapping with completely unique probes include: 1) Both the oligo-probe computational generation and its commercial synthesis is straightforward and much cheaper, 2) The number of oligo-probes required for the fingerprinting /mapping experiment is independent of the genome size and genome complexity, and 3) The informational-content output of each oligo-hybridization experiment is much higher - because, in opposite, the overwhelming majority of unique oligo-probe hybridizations to all DNA-clones in a library are negative. This approach makes possible a very natural and efficient probabilistic interpretation of the hybridization data, resulting in a strong robustness to experimental errors. In addition, the introduction of limited number of short perfect-match oligos, highly-specific to chosen known genes and/or functional sequences, can serve as a precise detection and localization data while reconstructing gene- and genome contigs - as well as to yield additional mapping and functional sequence data for the DNA-library investigated.

In computational simulations we defined a standard scenario with genome DNA-clones of length 40960 base pairs, in a genome of a length represented by 25 clone-lengths (totaling in approximately 1 million base pairs) and a coverage of 10. The simulated noise chosen included 20% false negatives (probability of a probe occurrence not to cause hybridization), and 5% false positives (probability of hybridization independently of probe occurrences). 500 specific oligo probes of length 8 (8-mers) were used. Based on the results of 1000 simulations in the standard scenario, the algorithm achieves probability of 96.1% of making no big errors whatsoever (defined as over-clone-size positioning errors between consecutive clones). The probability of making an over-clone size error between any two clones, not necessarily consecutive, was estimated at 7.5%. The average positioning error was 1740 base pairs (in clones of length 40960), and it can be further reduced using finer quantization in the algorithm, at some computational cost.

Real DNA is characterized by very-uneven base pairs distribution. As a result, most randomly chosen oligo probes occur only very-infrequently, whereas some probes occur extremely frequently. Therefore, practically, most probes will hybridize with hardly any DNA clones, or else - with almost all of them. Such hybridization fingerprint data may almost be completely useless. This is the scenario only, if the oligonucleotide hybridization is used for precise and complete sequencing. Indeed, with human DNA and 500 random 8-mers all constructed maps contained numerous errors. Variety of researchers have done vast amount of research investigation in this area during the past 20 years.

The solution to this problem lies in a judicious choice of probes – of their sequences, size, and total number to be used in oligo-hybridization experiment. The basic idea is that the probability of some oligo probe occurring on different regions of the same organism DNA is highly correlated and can be computationally investigated a priori. The first step is to use some already sequenced part of the organism's genome to estimate this probability for all possible probes of a certain length (this is achieved by the computer in a matter of minutes) that can be chosen from that particular genome sequence. Based on the obtained data, can be prepared a database of "good" oligo probes, which hybridize with about 50% of all DNA clones, and thus contain maximum overlap information. Our computational results with 500 pre-selected 8-mers were that 12% of all constructed maps were error-free.

While the result with pre-selected probes is already accommodating, it still needs improvement. The basic problem is, that while most of the pre-selected probes proved to be "good"-coverage probes for the physical mapping of a target genome, some experimentally proved to be very "bad" ones, and these where in fact numerous enough to confuse the algorithm analysis precision. In order to overcome this problem, we tried post-experimental selection of oligo-probes - by monitoring the total number of hybridizations and the performance of a particular oligo-probe and each of all the chosen oligo probes. In a hybridization experiment, about 250-300 post-selected probes remained performing well - out of the all 500 pre-selected ones. In the case of the Human DNA analysis experiment, 80% of the resulting maps were error-free. The fact that this result is better than the results with using large-size synthetic DNA probes, is explained by the fact that our criteria for "good" hybridization probes were more stringent than the range for random probes in synthetic long-size DNA-probes hybridization. Indeed, if post-selection is used also on long-size synthetic DNA - the error rate goes down to zero as well.

In a weeks-long experiment in 2003 using a super-computer, we analyzed the entire GenBank sequence data from all deposited genomes for the sequence-distribution of all possible oligos of the size from 8-mer long up to 24-mer long. Because we wanted to obtain information how the size of an oligo chosen corelate with its sequence representation among entire genomes data, regardless of the species it represents. Such results allow for choosing best oligo probes to be used in fingerprinting experiments regardless of the genome probed on (i.e. equally-important for all sequenced genomes and species). The result summary from this experiment is given below.

Table 2. Oligonucleotide n-mer analyses for the oligonucleotide distribution among GenBank database sequences (as of 2003). The first data gives the raw data. The interpretation of the number 754 in the "2" column of the 10-mer row is - that there were 754 distinct 10-mers that occurred in 2 distinct clusters in the hs.seq.all data set provided by GenBank. The "effective size" number is an estimate to the total number of n-mer occurrences within the total sequence data. Intuitively, this assumes that and n-mer may have only limited occurrence per gene, which is why the representation number in the cluster-column under cluster-size “1” increases with “n” increase from 8 to 24 – because cluster-size “1” calls for single representation among the genome data. So, among the Gen-Bank data from 2003: there are only 323 of all possible 10-mer oligo sequences represented uniquely (only once) among all known gene sequences deposed in GenBank, but 14,482,1196 of all possible 24-mer oligos are unique. On the other hand – there is no 8-mer oligo sequences that are uniquely existing in any of all deposited sequences in GenBank, so all the 8-mers exist in multiple occasions in all known genes, in fact, any 8-mer occur in more than 10 occasions among GenBank known genes.

Figure 9. Occurrence probability distribution of defined n-mer nucleotides in GenBank sequences – it shows the percentage of oligo distribution based on the effective size and demonstrates, that multiplicity of distribution matters. In the Table, it is shown, for example, that uniquely occurring 12-mers are as rare, almost as rare are multiply-occurring 24-mers.

A drawback of all of these approaches is that one must carefully choose, in advance, which type of sequences to probe. As a result, revisions to the arrays for correcting mistakes or incorporating new genomic informationare costly, requiring arrays to be redesigned and re-manufactured.

It is quite desirable to have a universal gene expression chip thatis applicable to all organisms, from bacteria to human, includingthose that lack complete cDNA libraries or whose genomes are notyetsequenced.

One way to obtain universality is to synthesize a combinatorial n-mer array containing all 4ⁿ possible oligos of length n, the key problem being to find avalue of n that is large enough to afford sufficient specificity,yet is small enough for practical fabrication and readout. Combinatorialn-mer arrays can be fabricated in a small number of simple stepsusing conventional solid phase synthesis chemistry and arraysof parallel fluid channels in perpendicular orientations to maskthe reagents as manufactured by Affymetrix company.

Until high-resolution, non-optical readout methods become practical, microarray densities, however, will be constrained by the optical diffraction limit. With this lower bound of ~0.28 µm on pixel size, n-mer arrays are limited to 8×10⁹ distinct spots per square inch, corresponding roughly to a 16-merarray on a 1" × 1" chip. Although it is possible to fabricatearrays with larger areas, we consider here arrays whose size (1-3-inch rectangle) is comparable to the current state of the art tofacilitate sensitivity comparisons. Therefore, we address thequestion of whether one can extract useful gene expression informationfrom combinatorial arrays of short (i.e., n

16)oligonucleotides.

Good practical approach to optimize the hybridization performance of the oligo-probes is using longer-size so-called “degenerated” oligos containing the nucleotide base inosine (I) that lack pairing (complementarity) specificity but improving the hybridization-duplex stability without influencing the hybridization specificity of the middle oligonucleotide “core”. For example, it is statistically known that 65,536 are all possible 8-mers (4⁸) to cover the entire genome, however it is practically impossible to achieve experimental conditions under which perfect-match discrimination results can be generated for all of the oligonucleotide sequences due to the specific experimental requirements. However reasonable compromises can be achieved in optimization of the hybridization conditions with an optimization in oligo size increase (up to 15-17-mers) and inosine “degeneration” of the oligonucleotide ends leaving the core sequence intact.

A similar strategy exploring long oligo probe approach is presented by Joseph DeRisi in a video we are recommending here below – for better visual explanation. We have to note, however, that the long-oligo approach described in the video is very-remotely similar and allows for very modest scale in gene identification by oligonucleotides – so much so, that on our opinion it is practically unsuitable for large-scale fingerprinting, nor for practical and precise gene identification. In its format, it can only serve as an example for an oligonucleotide-based genetic identification approach. ð Genome Sequencing for Pathogen Discovery - Joseph DeRisi (UCSF, HHMI) (https://www.youtube.com/watch?v=2LPfWuKBN_o).

DNA-clone specific signatures (finger-print codes) can be created for each clone of a solid-support arrayed library - by using even only 200 or more specific oligonucleotides or oligonucleotide-contained DNA- or RNA-fragments after their strong complementarity annealing to their corresponding target sequences via nucleotide hybridization (Figure 8). In a mirrored-type format, short oligonucleotides can be arrayed onto micro- or macro-array containing up to tens- or hundreds of thousands of short oligonucleotides of a size from 7- to 17-mers and hybridized with labeled DNA- or/and RNA pools originating from known, unknown and mixed organisms’ sources. Based on the hybridization data obtained at experimental conditions discriminating perfect-hybridization-match from a hybridization miss-match, followed by computational analysis, the previously unknown genetic-sequence fragments under investigation are successfully clustered and sorted according to their computationally-reconstructed sequence contigs (yielding un-interrupted sequence information). This strategy was named genetic Clone Fingerprinting and Recognition by Oligonucleotide Hybridization (CFROH). It consists of two steps: First, creating oligonucleotide-based clone-specific codes (fingerprints). Each gene-fragment fingerprint-code is generated as a result of the oligonucleotide hybridization process between a vast number of oligonucleotide probes with known sequences (as specially designed and chemically synthesized) and the gene fragments with unknown genetic content (generated from a biological source). Thereby the clone (DNA) oligo-fingerprints include: the total number and the corresponding primary sequences of all the positively annealed oligonucleotides (initially designed as within the range of 4-50 bases long), and the value (magnitude) of the hybridization intensities of the detected strong-annealing (hybridization) signal. Second, a comparison of this clone fingerprint to the analogous, computer generated “recoding” of the gene bank data bases, does generate a level of clone recognition. This is an oligonucleotide-code based clone-recognition or CFROH (or shortly - oligonucleotide fingerprinting) (Figure 10).

The original idea is the RECODING of the naturally-existing nucleotide-coded genetic information into an OLIGO-nucleotide based “code”. The basic reason for this approach procedure is the goal to reduce the total number of DNA- (RNA-) fragment specific signatures (naturally occurring as a nucleotide code) to a relatively much lower number of specific signatures (i.e. finger-prints) given by oligonucleotide hybridization data (occurrences & signal values), which as a result would demonstrate an ability to compare and recognize gene-fragments existing among unknown gene-fragment pools and genetic data-bases without the need for time- and resources-consuming precise sequencing of the entire genes and genomes. The critical importance of this strategic approach is most valuable for accessing genetic information within mixed libraries containing genetic clone-fragments generated from mixed population sources – usually serving large-scale detection-, diagnostic-, and monitoring purposes, especially in bio-pathogen detection and identification. So, for gene identification and analyses, we use oligonucleotide-based fingerprinting instead of complete gene sequencing.

The above approach provides both: significant decrease of all sequence-determination work, necessary for gene (-fragment) clone recognition and characterization, and possibility for complete gene-library characterization (DNA- or RNA-based) – each separated clone will obtain specific oligonucleotide-hybridization-based “fingerprint” code. As a result, it opens the possibility for a new level of gene marketing, that permits the sale of already separated and semi-characterized DNA-clones with completely or only partially knows genetic sequence without complete sequencing and deep resource spending.

Creating of a clone-specific signatures (code) by oligonucleotide hybridization.

Up to 31,000 cDNA-clones were arrayed onto a nylon membrane creating a chip (micro-array) of DNA fragments’ sequences. So, a cDNA-library of separated cDNA-clone fragments was created, which included a physically separated cDNA-clones each expressed in E. coli cells, their corresponding PCR-amplified inserts, and the inserts cDNA arrayed onto the nylon membrane (Figure 8).

At the first step of analyzing we obtained hybridization data, a DNA clone-dot concentration calibration and molarity-amount measurement on all chips (replicas) hybridized with approximately 200 oligos of defined and pre-selected sequences. Here are two important venues. First: The relations between molarities of all cDNA-clones dotted on each membrane-replica can be estimated in the hybridization step with an oligonucleotide, the complement of which is known to be presented in all PCR-products (so called mass-probe). This oligo usually is designed as a complement to the sequences placed between the PCR primer(s) and the foreign (cDNA) insert. If onto the membrane-filter are dotted controls with serial dilutions of the vector-DNA of known concentration, precise measurement and calculation of the molarities of all clone-dots is not a problem. This calculation and molarity measurement is very important for the future measurement and calculation of the hybridization intensities of all the cDNA-clones arrayed on the membrane – it allows to discriminate precisely the positively-hybridized cDNA-clones from semi-positive ones and the negative-hybridized. However, this type of measurement is not sufficient enough because each of the hundreds of oligonucleotide hybridization probes have not only different "affinity" to the cDNA-target than the mass-probe, but also because it is possible to hybridize to more than one target-site within the same cDNA-sequence – so it is necessary to measure the relation between the hybridization-intensity and the target-site molarity in each specific case of oligonucleotide hybridization event detected. Second: Additional strategy for target-site molarity measurement was developed. It includes arranging (on the membrane) repeatedly-spotted control dots of serial dilutions of specific complementary-oligonucleotide mixtures (containing non-self-complementary oligos) and also dots of serial dilutions of several very long (cosmid-DNA) fragments with known concentrations (pre-measured). The presumption is to have a control-representation of the complement-target of each possible oligonucleotide-prove used as a hybridization probe – so to be able to measure the target-molarity in each hybridization experiment and to eliminate the influence of the specific hybridization conditions and kinetics. Each control-target mixture should be printed at least as neighboring duplicated dots each time and deposited evenly across the array in a pattern similar to the one shown on the figure 11.

Figure 11. Scheme of control dots distribution onto the membrane array. The direction of the deposition of all repeats is shown with arrows.

The second step is a computational measurement of each hybridization result (i.e. each membrane-array image) (Figures 8 and 10). A custom program ("DOTS-program") or similar one is used for a measurement of each cDNA-dot onto the membrane, location of its precise coordinates among the pre-determined deposition matrix and transformation of the hybridization intensities into their numerical equivalent as a computer-generated signatures of the corresponding cDNA-clones (Figure 10). At this level the clone-specific code will be generated, consisting of: 1) The total number of all positively hybridized oligonucleotide probes, 2) the sequence of each oligonucleotide probe generating positive signal, and 3) the value of the detected hybridization signal as scaled among all clones and membranes – the specific measurement serving this purpose is the complement of the chosen oligo present in each particular clone (mass-probes) and in how many copies it is present that is a measure for the molarity relation between the oligonucleotide probe and it’s complementary target site onto the corresponding cDNA.

Figure 12. Comparison of clusters and clustering results obtained from 3 different clustering experiments – each containing different number of cDNA insert fragments from normalized (subtracted) cDNA library only (spot-attached onto DH-, DL-, BSM- and ZBS membrane arrays) and different number of oligonucleotide hybridization probes. The type of comparison and the computational clustering parameters are presented on the figure. Note, that at the left of the table, columns #1 and #2 containing comparison data from clustering experiments #1 (including BH - arrays) and #2 (including BH- and DL arrays) – are duplicated (top-to-bottom) for better data visualization.

The significance of similarity between a sequence and the oligo data-list was determined by the Algorithmic Significance method: For each cDNA sequence and a query oligomer list was estimated how many bits of information about the known sequence were revealed by the oligo-list hybridizations – where every bit of information implies a two-fold increase in significance value of the similar particular. Two parameters were considered for each query sequence – the top-score with a particular oligomer list, and the difference between the top-score and the second-highest score, for which were assigned the terms “absolute” and “relative” scores, respectively. A few inconsistencies among the results (different cDNA-sequences matching the same cDNA-clone) were resolved by taking the one with the highest absolute score.

At this point is hard to verify the real situation – as to how this oligo-based clustering correspond to the real cone similarity – are the clones of one cluster a real copy of the same- or a very-similar mRNA, i.e. are the clustered clones of one computed cluster are in fact members of genetically similar family of genes. To verify the accuracy of our clustering methodology, an additional approach was used to investigate the computational clustering data – 100 clusters were chosen to pick one cDNA-clone, a member of each one cluster and the cone cDNA-insert was hybridized as a probe against all another clones of the membrane (Figures 13, 14 and 15) in order to expose all really-similar clones out of the approximately 31,000 deposed onto the membrane and included in the same clustering experiment. Note, that the chip is prepared by the bottom-half of each chip being identical to the top part; so, each DNA fragment is attached twice on the chip (on the same row from the top to the bottom) in order to serve as additional positive control – i.e. each DNA-fragment signal should be detected as 2 replicas on the same chip and should produce equal computational output result.

The result did approve all logic of the entire investigation – as less oligonucleotide probes were used being included in the clustering experiments, as less precise the clustering result was.

So, (on Figure 13) when 107 oligos were used in generating hybridization data and in a clustering experiment, for some clusters all really-similar clones were clustered in the same cluster, but higher percent of clusters contained "mixed" clones. When 152 clones are used, some of the "mixed" clones become to be members of the same "correct" cluster; however, in all cases clusters contained more "ballast" clones, which were not detected to be of the same group in reality by the clone–>clones hybridization experiments. The above becomes as a logical result due to the low accuracy in the technical design for molarity target measurement and for hybridization intensities calculations used in this particular experiment (with 152 oligo-probes). Very successful become the clustering experiment including 217 oligonucleotide probes and 31,000 cDNA-clones. Here we demonstrate final experimental data for only 3 clusters and these data are really amazing (Figure 13 and figure 14): First, for a clone #1, recognized by partial sequencing to be an analog to human Guanidine-binding protein mRNA (Figure 13). Four clones, detected by self-hybridization to be members of a same cluster in reality, were clustered in the same cluster in several clustering experiments, but only if 217 oligonucleotide probes were used, the clustering is absolutely correct – only the same 4 clones are in one cluster. Second, for a clone #2, recognized to be an analog of thyroid hormone mRNA (Figure 13). Two clones, detected to be members of the same cluster in reality were clustered in different clusters, if 107 oligo-probes were used. Clustering analysis with 152 probes separate them in the same cluster, but contained many more "incorrectly" clustered clones. When 217 oligo-probes were used, the clustering analysis become again absolutely correct – both clones were separated within the same single cluster with no other clones in it!

Moreover, third clustering experiment (Figure 15, top; clustering with 107 oligo probes at threshold value of 100) identified eighteen cDNA clones clustered in 5 distinct clusters, but, when a single cDNA-clone was used as a similarity probe, it hybridized positively with 22 cDNA clones belonging to 5 different clusters establishing a high order of sequence similarities between all of them (Figure 15, bottom). Almost astonishingly, after a complete cDNA clone sequencing of single members of each of the 6 clusters and a sequence analysis, we found that all these DNA fragments did belong to different genes of a highly-similar family of genes. In this case the oligo-fingerprinting analysis was capable of highly precise clustering and gene discrimination – over-performing the capability of standard DNA-DNA hybridization techniques commonly used, with a performance close to the one of gene sequencing – as good as desired by the a priori design of our oligo-fingerprinting approach.

Figure 13. Comparison of the results from different clustering experiments and DNA insert"inserts hybridization. The clustering parameter values are shown on the figure. All DNA clones demonstrating highest positive hybridization signal (visible at the bottom) are marked with rhomb (·) within the table (at the top) and encircled on the images (at the bottom). As hybridization probe in the insert"inserts hybridization, was used a PCR-amplified DNA-insert of the clone located on the BSM1-array at row-45, column-9 (from row-8, column-2 within the MS15 PCR amplification plate). Labeling reaction was performed by the addition of 30 mCi of g-³³dATP into the PCR amplification reaction.

Figure 14. Fingerprinting clustering analyses with 107 and 217 oligo probes at threshold values from 60 to 100 (top of the figure) and DNA-fragment hybridization with probe identified as GenBank “X1785 Xenopus mRNA for thyroid hormone” against all DNA fragments attached to the chip (bottom of the figure).

Figure 15. Fingerprinting clustering analyses with 107 and 217 oligo probes at threshold values from 60 to 100 (top of the figure) and DNA-fragment hybridization with probe identified as GenBank “X03558 Human mRNA elongation factor” against all DNA fragments attached to the chip (bottom of the figure).

The results are very exciting, because they show, that entire cDNA clone library characterization and precise fingerprinting clustering is technically achievable and really applicable with relatively larger number of oligonucleotide probes. The precision of the clustering, for example, of only two DNA-clones out of 31,000 in a separated cluster, which was equal to the reality existing situation, is an amazing achievement. In one other case, the DNA-clone"attached-DNA clones hybridization exposed (detected) only 4 DNA-clones assumed to be really similar. (Figure 15, top). However clustering analysis showed two of them in the same cluster and the other 2 were in another 2 different clusters (Figure 15, bottom left). It was really amazing, that a stringent membrane wash (2h at 90°C) eliminated them as members of the same cluster (Figure 15, bottom right; note, that they are attached as top-to bottom replicas – i.e. exposed as 4 dots while being only 2 different DNA-clones – on the left and right on the chip image)! The additional 2 DNA-clones visualized on the left chip image (while absent on the image to the right) were very similar clones, but with slightly different sequences and this high-similarity was nevertheless discriminated by our fingerprinting approach proving to demonstrate precision higher than the standard clone-clone hybridization experiment..

Unknown DNA Sequence Recognition Based Upon Oligonucleotide Fingerprinting Results

The cDNA clone sequences were identified by comparing the list of oligomers’ sequences that were identified to occur in them (by the means of hybridization fingerprinting) against all known DNA sequences from GenBank. For each hybridization signature (fingerprint) consisting of specific number of hybridization intensities (signal values), a list was compiled – consisting of the corresponding oligo-sequences demonstrating only perfect-match hybridizations among all the 260 oligomer probes and exhibiting the highest hybridization intensities. The list of all the 260 oligomers was augmented by the additional 260 reverse complementary oligomers (because both strands of the PCR products were hybridized and the orientation of oligomers against the genetic data could not be resolved) – these served as additional sequence-data (and hybridization) verification since both reverse-complementary DNA strands are present in the labeled hybridization pool.

To identify the most frequently occurring cDNAs, the clones from the original Human brain cDNA library were clustered at high computational stringency (to maximize sequence-homogeneity of clones within clusters) and a single representative cDNA clone from the 100 largest clusters was selected for further investigation. The corresponding 260-oligomer lists were used as a database for searches using known DNA sequences as queries. A total of 195 searches involving gene sequences of average length of 2.5 kb were performed against the database consisting of the 260-oligomer lists.

Usage of degenerative oligo probes and capability of discriminating genes with extremely-high sequence similarities among large-scale DNA libraries containing highly abundant gene transcripts.

The “impossible dream” in the oligo-fingerprinting technology is achieving the ability to discriminate and identify unknown genes of families with relatively high abundance expressed at moderate levels from entire organisms’ libraries. For example, ability to detect specific immunoglobulin transcripts of unknown cell-clones between two different cDNA libraries obtained from two closely-related genetic lines of laboratory animals and, moreover, to discriminate between these highly-similar immunoglobulin gene sequences based solely upon the oligo-fingerprinting results.

For the purpose of testing our technology, we initially performed GenBank^® similarity search and analysis to identify a very-short regions within the immunoglobulin transcripts which would exhibit gene-specific (unique) sequences – aside of the otherwise extremely-high sequence similarity among all of the known immunoglobulins’ transcripts. With another words, at the current time, no data was known to exist for sequence variability of a chosen immunoglobulin gene or its mRNA gene family, nor any investigation was well documented as to the level of its transcription variability or the specific particular biological function served. In order to investigate the capabilities of different oligonucleotide designs and limited number of oligos to serve as discriminatory and finger-printing tool, we designed and probed relatively small number of degenerated oligos of sizes 13-mers, 15-mers, and 17-mers and hybridized them with labeled cDNA from 2 very-close mouse lines – NZW and NZBW – one consisting of healthy mice and another – exhibiting acute renal failure.

This experiment proved the exclusive capability of detecting single-nucleotide variation detection and single-nucleotide-based discrimination in both – gene expression pattern difference among very-closely related mammals and a single-type molecular gene profiling of an unknown genetic content. The oligonucleotide fingerprinting precision was so high, that it led to correction of the GenBank^® sequence data establishing a single-nucleotide deviation (or detected gene mutation) in the data submitted in GenBank^® by multiple investigators. The results are represented on the figures 16, 17 and 18 below.

The result on figure 16 demonstrates the ability of this technology to discover the presence of an unknown gene transcript based upon only a single-nucleotide diference - that in fact is the goal of this technology to achive.

Figure 16. Probing the hybridization conditions accuracy for single-base miss-match discrimination depending upon the size of the oligonucleotide used – 12-, 13- and 15-mer tested on 2 arrays hybridized with cDNA-pools obtained from 2 mouse lines – one healthy and another exhibiting acute renal failure. Our fingerprinting approach was so accurate, that it was able to detect (unintentionally by expectation) a single-nucleotide inaccuracy within the sequencing service initially performed for us by Genome System Inc – as shown above in red (correct perfect-match G) and blue (incorrect single-miss-match A) coloring. At the bottom are shown the GenBank similarity hits confirming the accuracy of the oligo-fingerprinting result shown above it)

Figure 17. Probing the hybridization conditions accuracy for single-base miss-match discrimination depending upon the size of the oligonucleotide used – 13-, 15- and 17-mer tested on 2 arrays hybridized with total cDNA-probes obtained from 2 mouse lines – one healthy and another exhibiting acute renal failure. Also probing the applicability of single- and double- degenerated-nucleotide for the detection of a possible-existing single-base gene variations and/or unknown gene detection.

On Figure 17 is demonstrated that the n596604-p152m17-1 single-degenerated probe does not detect noticeable signal difference among the two mice lines, but the n596604-p152m17-15 double-degenerated oligo probe was capable to detect the presence of unknown antibody gene-transcript within the sick mice population that is absent within the healthy mice population (note, that the signal in the Right-side image is significantly higher than on the Left-side). This is in fact an unintentional discovery possibly related to the pathogenic conditions among the particular mice population, which is an almost amazing real practical confirmation of the abilities of the oligo-fingerprinting approach pioneered by us.

Figure 18. Figure 18. Probing the hybridization conditions accuracy for single-base miss-match discrimination depending upon the size of the oligonucleotide used – 13-, 15- and 17-mer tested on 2 arrays hybridized with total cDNA-probes obtained from 2 mouse lines – one healthy and another exhibiting acute renal failure. Also probing the applicability of single- and double- degenerated-nucleotide for the detection of a priori unknown genes present in the samples. This is a different gene fragment than the one used on Figure 17

On Figure 18, in the middle (in Red, pasted in yellow) is shown that the original commercially sequenced gene does not hybridize positively with neither cDNA probe – suggesting for a mistake in the commercial sequence. Just above it, however, the single-base degenerated oligonucleotide exposes the presence of another perfect-match gene fragment that is present in both mice cDNA populations. Another perfect-match hit is detected (look above) by another double-base degenerated oligonucleotide that was identified to be highly abundant gene for the 28S ribosomal subunit.

The results shown on figures 17 and 18 clearly demonstrate the ability of the Oligonucleotide-Fingerprinting technology to identify even single-base genetic variability and to identify previously unknown gene transcripts, disease miss-alignments and the presence of pathogens. Further investigation of the condition detecting the presence of an unknown antibody within the sick mice and absent among the healthy ones (Figure 17) led to the discovery for the first time of a genetic defect among the mice line engineered to demonstrate renal failure. Thereby, the oligo-fingerprinting detection of the presence of unique and unknown antibody among the sick mice led us to the discovery of a pathogenic RNA-transcript, probably due to a fact that at some point a mice virus integrated its gene in a fatal place within the mouse genome that resulted in un-regulated and un-stoppable synthesis of viral RNA and viral protein (in the absence of the virus) that was triggering the mice immune system to attack it presuming there is alive virus present in its body – in a long term this possibly was leading to clogging the mice kidney with practically unnecessary antibodies against the viral protein resulting in renal failure and mice death.

We have devised and refined a technology involving strategy for entire gene‑libraries characterization by DNA-sequence information recoding from the 4-letter nucleotide code to a proprietary oligo‑nucleotide code - for the creation of oligonucleotide-hybridization based fingerprints of the probed gene fragments. Thereby, the Oligonucleotide Fingerprinting served for computational genetic clustering and identification of known and unknown genes, and establishing sequence similarity relations between genes from entire organism libraries with previously unknown genetic content.

By our low-scale oligo-fingerprinting approach, we proved the concept that our technology of analyzing the GenBank^® sequence data, proprietarily designing relatively limited number of short oligonucleotide sequences and experimental testing condition is capable of extremely precise detection of specific genes and gene transcripts (and their variations) present in any probe of any genetic background.

This technology demonstrates exceptional and unique ability to multiple detection of pathogens (known or novel), of any-type genetic deficiencies causing genetic diseases, detection of unknown antibodies expression as causes of autoimmune diseases, and the presence of potential biohazard agents in enclosed volumes of space (serving National biohazard-detection purposes) – this is ALL-AT-ONCE DETECTION ON OLIGONUCLEOTIDE MICROARRAY CHIPS.

This article is dedicated to support our efforts to obtain funding support for the technology refinement and for the preparation of compact oligonucleotide chips, and quick-processing equipment in order to ensure the production and market-establishment of this technology. This effort will serve for a revolutionary-wide spectrum of medical diagnostics tests – similar (but already practically-achieved long-ago in 1993-2003), to what the disgraced Theano’s company was promising falsely.

Moreover, this technology becomes extremely necessary to safeguard our National health- and bio-security for ever-increasing novel pathogens, research miss-happenings, and rogue bio-weaponizing – because it is universal, easily miniaturize-able and able to deliver rapid bio-detection in any enclosed volume (transport or cargo) within few hours.

We have also developed a research project to miniaturize the detection capabilities for applications using unmanned surveillance devices capable to deliver remotely pre-diagnostics data to be computationally processed at analytic centers and facilities.

For more information and funding discussions, please, contact Dr. Chris Dyanov at regontechnologies@gmail.com, regonmedical@gmail.com or at (773) 397-5496.