Small RNA (mirxon) signatures for determining biological states

Abstract

The present invention refers to a method for in vitro determining a change of a biological state of a subject or the difference of biological states by providing at least two mirxons from a reference sample, which represent a set of biological states, wherein said mirxons are small RNAs of specific sequence, further providing the same mirxons from a sample of the subject, quantifying the levels of one or more of said mirxons and correlating the levels of the mirxons of the reference sample with the levels of the mirxons of the subject, wherein any deviation of said levels reflects differences in said biological states, which are indicative for the change of the biological state to be determined, and normalizing the levels of the mirxons to the levels of one or more normalizers from the samples to a subset or all mirxons, or small RNAs, wherein the levels of said normalizers are most constant over the biological states.

Description

[0001] SMALL RNA (MIRXON) SIGNATURES FOR DETERMINING BIOLOGICAL STATES

[0002] FIELD OF THE INVENTION

[0003] The present invention provides methods for determining a change of a biological state using small ribonucleic acids (sRNAs), herein called mirxons, consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), wherein N can be any nucleotide; y and x are non-negative integers; the sum of x and y is not bigger than 6, and (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches, and combinations comprising sRNAs for use in the treatment of diseases and disorders.

[0004] BACKGROUND OF THE INVENTION

[0005] The reliable assessment of biological states opens a plethora of opportunities for the diagnosis and treatment of medical conditions and the evaluation of most different biological samples. The ubiquitous presence of polymeric RNA molecules in biological specimens makes them an essential component of a universally applicable detection tool.

[0006] Most RNA molecules with a length below 200 nucleotides (nt) are referred to as small RNAs and can be divided into different categories. Next to structural or housekeeping RNAs like transfer RNAs (tRNAs) and their fragments, ribosomal RNAs (rRNA), many functional small RNAs, there are regulatory small RNAs which can act in different manners. MicroRNAs (miRNAs) are endogenous single-stranded non-coding small RNAs with regulatory functions on gene expression. With an average length of 21- 22 nt most miRNAs originate from genes between other genes (intergenic) and some reside in introns of other genes. The latter miRNAs are also known as mirtrons — a term composed of miRNA and intron.

[0007] The main functions of miRNA include posttranscriptional regulation of gene expression and RNA silencing. Hence, miRNAs play a major role in the development, and differentiation of cells and their homeostasis. A considerable body of research, however, tells the other side of the story in which the dysregulation of miRNA results in disorders and diseases. The role of miRNAs is not only limited to intracellular processes but it has also been shown that extracellular miRNAs are part of physiological and pathological processes, and their expression profiles in body fluids are tightly related to various pathological conditions [1].

[0008] In view of their function as transcription-regulating factors via RNA silencing, small RNAs may play a crucial regulative role between genetic information and the resulting expressed proteins. Investigating and deciphering this interface is essential for a better understanding of biosynthesis and thus also of regulatory processes, which can be the reason for the development of various diseases. However, despite the role that small RNAs have in the biology of cells, increasing evidence points towards circulating small RNA profiling and the association of small RNAs with human disease, their diagnostic potential has not been fully realized.

[0009] W02021150990A1 discloses methods for constructing disease classifiers for evaluating subjects for biological conditions or diseases. Said classifiers being small RNAs (sRNAs, e.g. miRNA). W02021150990A1 discloses many small RNA species, however it does not disclose mirxons as described herein, which are functionally different from all the RNA species listed in W02021150990A1. Furthermore, in WO2021 150990A1 isomirs are grouped together by their seed sequences and either weighted in the model or aggregated.

[0010] WO2018136936A1 discloses a method for identifying or detecting small RNA (sRNA) predictors and a method for determining a condition of a subject by providing a biological sample, and identifying the presence or absence of sRNA predictors.

[0011] WO201 8076015A1 discloses a method for identifying a subject in need of therapeutic intervention to treat a disease or condition by isolating isomiRs or miRNAs from the subject, and characterizing the isomiRs or miRNAs and their presence or absence in the sample to identify a signature, wherein when the signature is indicative of a diagnosis of the disease then treatment of the subject is recommended.

[0012] US2007077582A1 discloses approaches for the detection and quantitation of short RNAs in a nucleic acid sample, for the determination of expression levels for two or more target short RNAs in a single reaction.

[0013] Hence there still remains a dire need to identify biomarkers able to determine the biological state of a subject or sample such as to determine or follow-up a disease status. Accurate molecular approaches for classifying or subtyping biological states such as complex diseases could lead to enormous breakthroughs in diagnosis and therapy, and to the next generation of patient analysis and care. SUMMARY OF THE INVENTION

[0014] The problem is solved by the embodiments of the present invention.

[0015] The present invention provides methods for evaluating subjects for one or more distinct biological states of interest, specifically for one or more disease subtypes.

[0016] According to the invention there is provided a method for in vitro determining a change of a biological state of a subject, comprising the steps

[0017] (a) providing at least two mirxons from a reference sample, which represent a set of biological states, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), wherein

[0018] N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0019] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches,

[0020] (b) providing the same mirxons of (a) from a sample of the subject,

[0021] (c) quantifying the levels of one or more of said mirxons and correlating the levels of the mirxons of (a) with the levels of the mirxons of (b), wherein any deviation of said levels reflects differences in said biological states, which are indicative for the change of the biological state to be determined, and

[0022] (d) normalizing the levels of the mirxons of (c) to the levels of one or more normalizers from the samples, or to a subset of mirxons or all mirxons of (a) or to other RNAs, specifically small RNAs, wherein the levels of said normalizers are most constant over the biological states.

[0023] According to a further embodiment of the invention, the reference sample is from one or more subjects of the same species, specifically said sample is from healthy subjects, or subjects of one or more specific biological states.

[0024] A further embodiment of the invention provides a method for in vitro monitoring the treatment of a subject, comprising the steps

[0025] (a) providing at least two mirxons from a reference sample, which represent a set of biological states, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), wherein

[0026] N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0027] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches,

[0028] (b) providing the same mirxons of (a) from a sample of the subject,

[0029] (c) quantifying the levels of one or more of said mirxons and correlating the levels of the mirxons of (a) with the levels of the mirxons of (b), wherein any deviation of said levels reflects differences in said biological states, which are indicative for the change of the biological state to be determined, and

[0030] (d) normalizing the levels of the mirxons of (c) to the levels of one or more normalizers from the samples, or to a subset of mirxons or all mirxons of (a) or to other RNAs, wherein the levels of said normalizers are most constant over the biological states.

[0031] According to a specific embodiment of the invention, the normalized levels are analyzed with a mathematical or logical method to provide information on the probability of a biological state, condition, or change of the subject.

[0032] According to a specific embodiment of the invention, the normalized levels are combined with further RNA levels, or biological, clinical, environmental, experimental, or biographical data.

[0033] According to a specific embodiment of the invention, the normalizers are selected from the group consisting of RNAs, small RNAs, functional RNAs, transfer RNAs or their fragments, circular RNAs, large RNAs, coding RNAs, non-coding RNAs, peptides, proteins, DNA sequences, metabolites, lipids, carbohydrates, internal controls, cell counts, tissue weights, and any combination or subset thereof.

[0034] According to a further embodiment of the invention, the difference in the levels of the mirxons is determined by endpoint, quantitative, or digital PCR, miRCURY assays, sequencing, microarrays, Luminex nucleic acid assays, Nanostring assays, or other hybridization-based techniques.

[0035] According to a specific embodiment of the invention, the sample is selected from the group of blood, extracellular vesicle, tissue, cell, cell cluster, organoid, spheroid, cell aggregate, ascites, pleural effusion, stool, urine, sperm, saliva, menstrual blood, bile, bone marrow aspirate, breast milk, cerebrospinal fluid, plasma, serum, sweat, sputum, synovial fluid, lymph, oral swab, nasal swap, oropharyngeal swap, nasopharyngeal swap, vaginal swab, urethral swab, cervical swab or lavage, smear of a surface, tissue, and body cavity or cell free supernatant of in vitro grown tissues, cells or cell aggregates.

[0036] A further embodiment of the invention provides a method for monitoring a subject, specifically for monitoring the development, progression, or change of a biological state.

[0037] According to a specific embodiment of the invention, the biological states are selected from the group consisting of biological, physical, chemical, clinical, developmental and pathological states, specifically malignant disease, cardiovascular disease, endocrine disorder and disease, immunological disease, mental disorder, psychological disorder and disease, infection, malnutrition, intoxication, injury, cellular sternness, cellular developmental state, cellular degree of differentiation, or a combination thereof.

[0038] According to a specific embodiment of the invention, the mirxons comprise mismatches at the C nucleotides of the (CGG)e (SEQ ID NO: 3) nucleotide sequence.

[0039] According to a specific embodiment of the invention, the mirxons comprise mismatches at the G nucleotides of the (CCG)e (SEQ ID NO: 4) nucleotide sequence.

[0040] According to a specific embodiment of the invention, the mirxons comprise mismatches which are one-base deletions or insertions.

[0041] A further embodiment of the invention provides a library comprising ten or more mirxons, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx, (SEQ ID NO: 2) wherein N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0042] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

[0043] According to a further embodiment of the invention, the library is constructed by isolating and selecting one or more mirxons from one or more samples.

[0044] A further embodiment of the invention provides a reference sample library comprising data sets of at least ten mirxon levels, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)eNy(SEQ ID NO: 1) or Ny(CCG)eNx (SEQ ID NO: 2), wherein N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0045] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

[0046] A further embodiment of the invention provides a composition comprising one or more mirxons, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), wherein N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0047] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

[0048] A further embodiment of the invention provides a pharmaceutical composition comprising one or more mirxons, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), wherein N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

[0049] According to a specific embodiment of the invention, the bases or backbone of said mirxons in said composition and pharmaceutical composition are replaced by molecules not naturally occurring in RNA molecules, modified ribose sugars, or amide linkages. Said molecules are selected from the group of, but not limited to, deoxyribonucleotide bases, artificial bases, locked nucleic acid (LNA) bases, or equivalent molecules.

[0050] According to a specific embodiment of the invention said composition and pharmaceutical composition may be used in the treatment of a disease or disorder, specifically selected from the group consisting of malignant disease, cardiovascular disease, endocrine disorder and disease, immunological disease, mental disorder, psychological disorder and disease, infection, malnutrition, intoxication, and injury.

[0051] FIGURES

[0052] Figure 1 : Primer sequences and melting temperatures for RT-qPCR.

[0053] Figure 2: Sequence logos of mirxons. Figure 3: Overlap of newly predicted mirxons from OVARY and ENCODE up to five and six mismatches.

[0054] Figure 4: miRNA sequence analysis for three-base repetitive sequences, origin and origin exon length. Y-axis panel B: Y axis: Percent

[0055] Figure 5: Sequence logos or mirxons up to four mismatches.

[0056] Figure 6: miRNA expression and expression variability analysis. Y-axis panel B: Log2 ratio mirxons to all miRNAs. Y-axis panel D: Log2 ratio mirxons to all miRNAs.

[0057] Figure 7: Genome plots of the genomic loci of three mirxons which were experimentally verified by RT-qPCR.

[0058] Figure 8: RT-qPCR results for the two mirxons PCR-systems in four cancer cell lines upon silencing. Y-axes: Log2 expression (arbitrary unit).

[0059] Figure 9: Impact of mirxon silencing (using an anti-miR) on gene expression of four cell lines.

[0060] Figure 10: Expression of five significant genes (RT-qPCR) upon silencing. Y- axes: Log2 expression (arbitrary unit).

[0061] Figure 11 : Genomic loci of five significant deregulated genes upon silencing of the (CGG)e (SEQ ID NO: 3) mirxon (using (GGC)e (SEQ ID NO: 4) anti-miR.

[0062] Figure 12: Log2-ratios of all mirxons up to five mismatches (“Mirxons”) and all miRBase miRNAs (“MiRBase”) to all miRNA reads, split into samples derived from Dicerpositive (293T, left) and Dicer-negative (Dicer KO, right) cells.

[0063] Figure 13: Splines of the hazard ratio of a Cox regression model correlated to the ratio of mirxon derived reads related to all miRNA associated reads and correlation of the expression of the (GGA) and (GGGA) mirxons.

[0064] Figure 14: Impact of mirxon expression on outcomes in cancer patients.

[0065] Figure 15: Boxplots showing the difference of gene-signatures derived from the mirxon silencing experiment. The y axis: Log2 expression (arbitrary unit).

[0066] Figure 16: Log2 ratios of mirxons associated reads to all miRNA associated reads in pure tumor cells of high grade serous ovarian cancer subdivided in derived from ascites (AS) or solid tumors (PM) and the respective peritoneal tumor spread (i.e. miliary (Mi) and non-miliary (nMi)). The y axis: Log2 expression (arbitrary unit).

[0067] Figure 17: Log2 ratios of mirxon associated reads in extracellular vesicles. Y-axis panel A and B: Log2 ratio mirxons (with mismatches indicated at the x-axis) to all miRNAs. Panel C y-axes from left to right: Ratio, Concentration, Motility, Morphology. Figure 18: Distributions of mirxon ratios from cell-free cell culture supernatant from HNSCC-derived and healthy-derived cells and difference of the ratio between both groups. Right panel: the significant difference of the ratios of MM0_4 (mirxons with zero to four mismatches) between benign and malignant (i.e. HNSCC) is shown (0 refers to benign, 1 refers to malign). Y-axis left plot: Log2 ratio mirxons (with mismatches indicated at the x-axis) to all miRNAs.

[0068] Figure 19: Distributions of mirxon ratios from extracellular vesicles (EVs) of endometrial fluids (EF) of endometrial fluid and difference of the ratio between non- implantative and implantative endometrial fluid. Right side: the significant differences between implantative and non-implantative endometrium of the ratios: mirxons up to six mismatches divided by all miRNA reads (MM0_6), are shown. Y-axis left plot: Log2 ratio mirxons (with mismatches indicated at the x-axis) to all miRNAs.

[0069] Figure 20: Distributions of mirxon ratios from extracellular vesicles (EVs) of colorectal cancer (CRC) patients and controls and difference of the ratio between healthy controls and CRC-patients. Right side: the significant differences between healthy and colorectal cancer patients of the ratios: mirxons up to four mismatches divided by all miRNA reads (MM0_4) are shown. Y-axis left plot: Log2 ratio mirxons (with mismatches indicated at the x-axis) to all miRNAs

[0070] Figure 21 : A, the boxplots at time points 1 and 2 show the predictor values of the 11 -mirxon model of the 70 healthy controls and the 90 CRC patients, respectively, proving an excellent discriminative power. The line with open circles represents the 11- mirxon model values over the 13 follow-up time points (left y-axis). The line with the open triangles represents the tumor size at follow-up examinations (right y-axis). B, the same is seen from the best discriminative miRBase miRNA hsa-miR-320c, line with closed circles, and tumor sizes, line with closed squares.

[0071] Figure 22: The best discriminating miRBase miRNA hsa-miR-4733-3p and the best discriminating mirxon atlasMiR-106609-3p, a mirxon with three mismatches, are shown. The Y-axes show the relative Iog2 expression (arbitrary unit).

[0072] Figure 23: Three tissues with the best age-dependent miRBase miRNA mmu- miR-29c-3p and the best age-correlated mirxon, mirxon-1895-3p are shown. The Y-axes of Figure 23 show the relative Iog2 expression (arbitrary unit).

[0073] Figure 24: Log2 fold-changes of CGI (CpG-island) methylation beta-values correlated to their corresponding mirxon expression values. Figure 25: Log2 fold-changes of GB (gene body) methylation beta-values correlated to their corresponding mirxon expression values.

[0074] Figure 26: Log2 fold-changes of CGI and GB methylation beta-values correlated to their corresponding mirxon expression values according their mean methylation betavalues (X-axis).

[0075] Figure 27: Potential of repetitive sequences and miRNAs for triplex helix formation. Panel A y-axis: Frequency.

[0076] DETAILED DESCRIPTION

[0077] Unless indicated or defined otherwise, all terms used herein have their usual meaning in the art, which will be clear to the skilled person.

[0078] The terms “comprise”, “contain”, “have” and “include” as used herein can be used synonymously and shall be understood as an open definition, allowing further members or parts or elements. “Consisting” is considered as a closest definition without further elements of the consisting definition feature. Thus “comprising” is broader and contains the “consisting” definition.

[0079] The term “about” as used herein refers to the same value or a value differing by + / - 5 % of the given value.

[0080] As used herein and in the claims, the singular form, for example “a”, “an” and “the” includes the plural, unless the context clearly dictates otherwise.

[0081] The terms “individual” or “subject” are meant to broadly include any multicellular eukaryotes, specifically animals, specifically mammals, specifically humans.

[0082] As used herein, a “healthy subject” or “healthy individual” refers to any control subject or any population of control subjects who have no diagnosable disease or disorder, or are not suffering from a specific disease or disorder which is to be determined as described herein.

[0083] As used herein, the term “most constant” refers to the distribution of a data set wherein the data points have a relatively narrow distribution, specifically a standard deviation not bigger than 5 % of the mean, specifically 0.0, 0.25, 0.5, 0.75, 1.0, 1.25, 1.5, 1.75, 2.0, 2.25, 2.5, 2.75, 3.0, 3.25, 3.5, 3.75, 4.0, 4.25, 4.5, 4.75, or 5.0 %. In a preferred embodiment the distribution has a skewness of -2 to 2, specifically -1 to 1 , specifically -2.0, -1.5, -1.0, -0.5, 0.0 (i.e. symmetric distribution), 0.5, 1.0, 1.5, or 2.0. The present invention provides an in vitro method for determining a distinct biological state of interest, which may also be a disease or a disease subtype. A disease subtype may be a collection of biological states that manifest with similar disease symptoms, but which may involve distinct sRNAs and / or which may require different modes of treatment. The present invention also provides a method for determining the change of a biological state of interest. The presence, absence, or abundance of one or more mirxons (e.g. expression level) across a sample set can be predictive of a biological state or the change of said biological state, or is predictive for a disease progression or response to treatment. The use of the mirxons as described herein may accurately classify biological states that may be present with similar symptoms or pathologies, including early stages of a disease. With the method described herein, the presence or change of at least 2, 3, 4, 5, 6, 7, 8, 9, 10, or more biological states can be determined, even if they share a common phenotype or pathology.

[0084] As used herein, the term “biological state” refers to the present state or condition of a cell, tissue, or organism, wherein said state is determined by internal and / or external influences such as, but not limited to, biological processes, genetics, epigenetics, immunological processes, developmental processes, pathological condition, physiological processes, physical stimulus, chemical stimulus, metabolism, or a combination thereof. The biological state can manifest with certain clinical symptoms or have a certain pathologic annotation.

[0085] Examples of “pathological states” can be, but are not limited to, malignant disease, cardiovascular disease, endocrine disorder / disease, immunological disease, mental disorder, psychological disorder / disease, infection, malnutrition, intoxication, injury, infertility, endometriosis, gynecological disorders or any combinations thereof. A disease is a pathological process, whereas a disorder is characterized by functional impairment and a disruption to the body’s normal function and structure, however, both terms can also be used interchangeably herein.

[0086] Exemplary phenotypes or pathologies that may define the biological state(s) of interest may include, but are not limited to one or more of cancerous malignancies; malignancy invasion; dementia; neurodegeneration; demyelination; anxiety, depression, headache or fatigue; insomnia; chronic tissue inflammation; vasculitis; vascular permeability; irritable bowel syndrome, , and / or weight loss; muscle or joint pain or fatigue; gastrointestinal permeability; muscle atrophy; autoimmunity; tissue fibrosis; disorders of physical, mental, or social development; lysosomal storage abnormality; glycogen accumulation; uncontrolled cell proliferation; cell or tissue necrosis or apoptosis; fatty liver or hepatitis; chronic kidney disease; neutrophilia or neutropenia; bone remodeling abnormalities including abnormal osteogenesis or bone resorption; insulin resistance; hypertension; hypotension; vasoconstriction; pathological angiogenesis or lymphangiogenesis; hypercholesterolemia; metabolic disease or obesity; coronary artery disease; congestive heart failure; drug response or drug toxicity; among others.

[0087] The term “change of a biological state” refers to any deviation of said biological state from its initial point of determination in a specific frame of time.

[0088] The term “difference in biological states” refers to quantification of the comparison of one or more biological states, specifically, comparison of 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10,

[0089] 11 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 , 32, 33,

[0090] 34, 35, 36, 37, 38, 39, 40, 41 , 42, 43, 44, 45, 46, 47, 48, 49, 50, 51 , 52, 53, 54, 55, 56,

[0091] 57, 58, 59, 60, 61 , 62, 63, 64, 65, 66, 67, 68, 69, 70, 71 , 72, 73, 74, 75, 76, 77, 78, 79,

[0092] 80, 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91 , 92, 93, 94, 95, 96, 97, 98, 99, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220, 230, 240, 250, 260, 270, 280,

[0093] 290, 300, 310, 320, 330, 340, 350, 360, 370, 380, 390, 400, 410, 420, 430, 440, 450,

[0094] 460, 470, 480, 490, 500, 510, 520, 530, 540, 550, 560, 570, 580, 590, 600, 610, 620,

[0095] 630, 640, 650, 660, 670, 680, 690, 700, 710, 720, 730, 740, 750, 760, 770, 780, 790,

[0096] 800, 810, 820, 830, 840, 850, 860, 870, 880, 890, 900, 910, 920, 930, 940, 950, 960,

[0097] 970, 980, 990, or 1000 biological states

[0098] The term “set of biological states” refers to a collection of at least one biological state, specifically at least 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 , 32, 33, 34, 35, 36, 37, 38, 39, 40, 41 , 42,

[0099] 43, 44, 45, 46, 47, 48, 49, 50, 51 , 52, 53, 54, 55, 56, 57, 58, 59, 60, 61 , 62, 63, 64, 65,

[0100] 66, 67, 68, 69, 70, 71 , 72, 73, 74, 75, 76, 77, 78, 79, 80, 81 , 82, 83, 84, 85, 86, 87, 88,

[0101] 89, 90, 91 , 92, 93, 94, 95, 96, 97, 98, 99, 100, 110, 120, 130, 140, 150, 160, 170, 180,

[0102] 190, 200, 210, 220, 230, 240, 250, 260, 270, 280, 290, 300, 310, 320, 330, 340, 350,

[0103] 360, 370, 380, 390, 400, 410, 420, 430, 440, 450, 460, 470, 480, 490, 500, 510, 520,

[0104] 530, 540, 550, 560, 570, 580, 590, 600, 610, 620, 630, 640, 650, 660, 670, 680, 690,

[0105] 700, 710, 720, 730, 740, 750, 760, 770, 780, 790, 800, 810, 820, 830, 840, 850, 860, 870, 880, 890, 900, 910, 920, 930, 940, 950, 960, 970, 980, 990, or 1000 sets, specifically in the sets are in the range of 1-1000.

[0106] The term “in vitro method” as used herein refers to methods performed outside of a living organism. Specifically, it refers to methods performed on samples such as isolated tissues, organs, or cells, preferably blood, serum, or plasma. Such an in vitro method is therefore not performed on the living organism; it is particularly not performed on humans.

[0107] As used herein, the terms “nucleotide” or “base” can be used interchangeably and shall be understood as the individual units of nucleic acid polymers.

[0108] As used herein, the sense of nucleotide sequences, unless stated otherwise, is read from the 5’ end to the 3’ end.

[0109] As used herein, all nucleotide sequences described in this work can also be given with the nucleobase thymine (T) instead of the uracil (U), regardless whether a ribonucleic acid (RNA) or a deoxyribonucleic acid (DNA) nucleobase sequence is meant.

[0110] As used herein, “small RNA” relates to RNA molecules with a number of nucleic acid bases in the range of 2-200, specifically at least 10, specifically 18, 19, 20, 21 , 22, 23, or 24, specifically a least 30, specifically at least 40, specifically at least 50, specifically at least 100, specifically at least 150, specifically 200. Examples of small RNAs, specifically used as “normalizers” herein, include 5.8S ribosomal RNA (rRNA), 5S rRNA, tRNAs, miRNAs, siRNAs, snoRNAs, piwi-interacting piRNAs, tRNA-derived small RNAs (tsRNAs), and small rDNA-derived RNAs (srRNAs).

[0111] The term “mirxon” is composed of the word miRNA and exon, comparable to mirtron (miRNA and intron). As used herein, the term “mirxon” refers to an RNA molecule wherein the nucleotide sequence consists of Nx(CGG)eNy (SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), wherein N can be any nucleotide, y and x are non-negative integers, specifically 0, 1 , 2, 3, 4, 5, or 6, and the sum of x and y is not bigger than 6, specifically 0, 1 , 2, 3, 4, 5, or 6.

[0112] In a specific embodiment, Nxand Nyare referred to as flanking sequences. In some embodiments, flanking sequences may be used to direct the specificity of mirxons, specifically flanking sequences may be used to increase specificity.

[0113] Mirxons have a typical length of 18-24 nucleotides, specifically 18, 19, 20, 21 , 22, 23, or 24 nucleotides. The nucleotide sequence can show up to 6 mismatches at the C and G nucleotides of the (CGG)e (SEQ ID NO: 3) and (CCG)e (SEQ ID NO: 4) nucleotide repeats, respectively. Specifically, the mismatches within the trinucleotide repeats include, but are not limited to, AGG, TGG, GGG, CCA, CCT, and CCC, or combinations thereof.

[0114] Mismatches may also be one base deletions or insertions.

[0115] Within the (CGG)e (SEQ ID NO: 3) nucleotide sequence, the mismatches may preferably be at the C nucleotide position. Within the (CCG)e (SEQ ID NO: 4) nucleotide sequence, the mismatches may preferably be at the G nucleotide position. Specifically, mirxons, as used herein, are differentiated by using the mismatches and one or both flanking sequences, and the individualized sequences are used as separate predictors to differentiate said mirxons.

[0116] The term “first exons” refers to the the first exons (5’) of a gene, specifically, the term “first exons” refers to the first 1-10, 1-9, 1 -8, 1 -7, 1 -6, 1-5, 1-3, or 1 -2 exons or the first exon located in the 5’ region of a gene.

[0117] Mirxons may further be characterized by predicted stem-loop structures (pre- miRNAs) nearly always located in one of the first exons or promoter regions of other (protein-coding) genes, i.e. transcript variants thereof, (and due to their high CG-content, CpG-islands) and that these first exons are median 315 bp longer compared to first exons without mirxons.

[0118] Mirxons may further be characterized by their absence in the RNA-induced silencing complex (RISC), however, mirxons are bound to DNA directly. Therefore, some small RNA analysis methods would not even detect mirxons (e.g. methods using the LEXOGEN® TraPR Small RNA Isolation Kit).

[0119] The function of mirxons differs from any other class of small RNAs. Specifically, the localization of mirxons is in promoter regions or first exons of genes, specifically of protein-coding genes, and often in CpG-islands; mirxons directly bind to DNA at their place of origin. In specific embodiments mirxons directly bind to other regions on the DNA with similar sequences; mirxons are not present in the RISC complex;

[0120] DICER (a multidomain endonuclease that belongs to the fourth class of type III ribonuclease (RNase III) enzymes) is not involved during maturation of mirxons; and mirxons positively, or in some embodiments negatively, regulate gene expression at the transcriptional level, therefore mirxons have a direct impact on gene expression if silenced or overexpressed. In some embodiments, “at least two mirxons” refers to 2- 6000 mirxons, specifically 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 , 32, 33, 34, 35, 36, 37, 38, 39, 40, 41 , 42, 43,

[0121] 44, 45, 46, 47, 48, 49, 50, 51 , 52, 53, 54, 55, 56, 57, 58, 59, 60, 61 , 62, 63, 64, 65, 66,

[0122] 67, 68, 69, 70, 71 , 72, 73, 74, 75, 76, 77, 78, 79, 80, 81 , 82, 83, 84, 85, 86, 87, 88, 89,

[0123] 90, 91 , 92, 93, 94, 95, 96, 97, 98, 99, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190,

[0124] 200, 210, 220, 230, 240, 250, 260, 270, 280, 290, 300, 310, 320, 330, 340, 350, 360,

[0125] 370, 380, 390, 400, 410, 420, 430, 440, 450, 460, 470, 480, 490, 500, 510, 520, 530,

[0126] 540, 550, 560, 570, 580, 590, 600, 610, 620, 630, 640, 650, 660, 670, 680, 690, 700,

[0127] 710, 720, 730, 740, 750, 760, 770, 780, 790, 800, 810, 820, 830, 840, 850, 860, 870,

[0128] 880, 890, 900, 910, 920, 930, 940, 950, 960, 970, 980, 990, 1000, 1100, 1200, 1300,

[0129] 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 2600, 2700,

[0130] 2800, 2900, 3000, 3100, 3200, 3300, 3400, 3500, 3600, 3700, 3800, 3900, 4000, 4100,

[0131] 4200, 4300, 4400, 4500, 4600, 4700, 4800, 4900, 5000, 5100, 5200, 5300, 5400, 5500,

[0132] 5600, 5700, 5800, 5900 or 6000 mirxons.

[0133] In some embodiments, “one or more mirxons” refers to 1-6000 mirxons, specifically 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23,

[0134] 24, 25, 26, 27, 28, 29, 30, 31 , 32, 33, 34, 35, 36, 37, 38, 39, 40, 41 , 42, 43, 44, 45, 46,

[0135] 47, 48, 49, 50, 51 , 52, 53, 54, 55, 56, 57, 58, 59, 60, 61 , 62, 63, 64, 65, 66, 67, 68, 69,

[0136] 70, 71 , 72, 73, 74, 75, 76, 77, 78, 79, 80, 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91 , 92,

[0137] 93, 94, 95, 96, 97, 98, 99, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220, 230, 240, 250, 260, 270, 280, 290, 300, 310, 320, 330, 340, 350, 360, 370, 380,

[0138] 390, 400, 410, 420, 430, 440, 450, 460, 470, 480, 490, 500, 510, 520, 530, 540, 550,

[0139] 560, 570, 580, 590, 600, 610, 620, 630, 640, 650, 660, 670, 680, 690, 700, 710, 720,

[0140] 730, 740, 750, 760, 770, 780, 790, 800, 810, 820, 830, 840, 850, 860, 870, 880, 890,

[0141] 900, 910, 920, 930, 940, 950, 960, 970, 980, 990, 1000, 1100, 1200, 1300, 1400, 1500,

[0142] 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 2800, 2900,

[0143] 3000, 3100, 3200, 3300, 3400, 3500, 3600, 3700, 3800, 3900, 4000, 4100, 4200, 4300,

[0144] 4400, 4500, 4600, 4700, 4800, 4900, 5000, 5100, 5200, 5300, 5400, 5500, 5600, 5700,

[0145] 5800, 5900 or 6000 mirxons.

[0146] As used herein, the term “mirxon level” refers to the expression of the said mirxon. In some embodiments, the mirxon levels are determined by means of, but not limited to, endpoint, quantitative, or digital PCR, miRCURY assays, sequencing, single-molecule sequencing, mass spectrometry, microarrays, Luminex nucleic acid assays (e.g. ProcartaPlex™ Immunoassays), nanostring assays, or other hybridization techniques, or combinations thereof.

[0147] In some embodiments of the present invention, the levels of the quantified mirxons may be normalized to the levels of all mirxons from a reference sample, wherein the one or more of the quantified mirxons may also be included in the mirxons of the reference samples.

[0148] In an alternative embodiment, the levels of the quantified mirxons may be normalized to the levels of all mirxons from a reference sample, wherein the one or more of the quantified mirxons are excluded from the mirxons of the reference samples.

[0149] The RNAs, specifically used as “normalizers” described herein, can be doublestranded RNA or single-stranded RNA. The RNA can be circular RNA. The RNA can be, but is not limited to a bacterial (e.g., 5S, 16S rRNA, or 23S rRNA) or eukaryotic rRNAs (e.g., 5S rRNA, 5.8S rRNA, 18S rRNA, or 28s rRNA).

[0150] Functional RNAs are not translated into protein but have a biological role directly.

[0151] As used herein, “circular RNA” (circRNA) relates to a single stranded RNA molecule whose ends are covalently bonded together, thus forming closed loops.

[0152] As used herein, “large RNA” relates to RNA molecules with a length of at least 200 nucleic acid bases.

[0153] As used herein, “coding RNA” relates to an RNA molecule that is translated into a polypeptide or protein.

[0154] As used herein, “non-coding RNA” (ncRNA) relates to an RNA molecule that is not translated into a polypeptide or protein. Examples of a non-coding RNAs (ncRNA) that are not translated into a protein include transfer RNAs (tRNAs) and ribosomal RNAs (rRNAs), as well as small non-coding RNAs such as microRNA (miRNA), small interfering RNA (siRNA), Piwi-interacting RNA (piRNA), small nucleolar RNA (snoRNA), small nuclear RNA (snRNA), extracellular RNA (exRNA), small Cajal body-specific RNAs (scaRNAs), and long ncRNAs such as Xist and HOTAIR

[0155] As used herein, “small nuclear RNA” (snRNA) relates to a class of small RNA molecules that are found within the nucleus of eukaryotic cells.

[0156] In some embodiments, the RNA contains one or more post-transcriptional modifications comprising, but not limited to, RNA methylation, phosphorylation, etc.

[0157] In some embodiments, the polynucleotide comprises one or more “artificial nucleotide” analogues. In some instances, the artificial nucleotide analogues comprise modifications at one or more of ribose moiety, phosphate moiety, nucleoside moiety, or a combination thereof. In some embodiments, one or more of the artificial nucleotide analogues are resistant toward nucleases such as ribonuclease such as RNase when compared to natural polynucleotides. Examples of artificial nucleotides may include, but are not limited to one or more, locked nucleic acid (LNA), peptide nucleic acid (PNA), glycol nucleic acid (GNA), threose nucleic acid (TNA) and hexitol nucleic acids (HNA).

[0158] As used herein, the term “peptide” refers to any unbranched biological polymer that is composed of amino acids linked by peptide bonds.

[0159] As used herein, the term “DNA” refers to any polymer consisting of natural, artificial, or a combination thereof, nucleotides. In some embodiments, the DNA contains one or more modifications, such as, but not limited to, methylation, phosphorylation etc.

[0160] As used herein, a “metabolite” is any substance that is a product of metabolic action or that is involved in a metabolic process including e.g., each compound input into a metabolic reaction, each compound produced by a metabolic reaction, each enzyme associated with a metabolic reaction, and each cofactor associated with a metabolic reaction.

[0161] In some embodiments the “internal control” includes, but is not limited to, small RNAs, 18S rRNA, hsa-miR-191-5p, 5S rRNA, tRNAs, snRNAs, or combinations thereof.

[0162] As used herein, a “sample of the subject” can be selected from, but is not limited to, the group consisting of blood, extracellular vesicle, tissue, cell, organoid, spheroid, cell cluster, cell aggregate (the last six either from in vivo or in vitro cultured), ascites, pleural effusion, stool, urine, sperm, saliva, menstrual blood, bile, bone marrow aspirate, breast milk, endometrial fluid, cerebrospinal fluid, plasma, serum, sweat, sputum, synovial fluid, lymph, oral swab, nasal swap, oropharyngeal swap, nasopharyngeal swap, vaginal swab, urethral swab, cervical swab or lavage, smear of a surface, and body cavity or cell free supernatant of in vitro grown tissues, cells or cell aggregates.

[0163] As used herein, “reference sample” refers to a sample of known composition and / or having or known to have or lack specific properties (e.g., known nucleic acid variant(s), known cellular origin, known tumor fraction, known coverage, or the like) that is analyzed along with or compared to the samples from a subject in order to evaluate biological states.

[0164] As used herein, “correlating levels” refers to the statistical evaluation of the correlation or dependence between the levels of two sets of mirxons. As used herein, “normalizing the levels” refers to aligning the quantified levels of mirxons to a common scale by means of scaling, Z-score normalization, linear normalization, “Max-Min” normalization, clipping normalization, standard deviation normalization, cyclic loess normalization, quantile normalization (e.g. Aquantile, Gquantile, Rquantile or Tquantile), variance-stabilizing transformation (all methods for normalizing a larger number of mirxons), or normalizing to a set of normalizers (for individual or a set of mirxons).

[0165] As used herein, the term “normalizer” refers to any quantifiable object that can be used to normalize the levels. Normalizers can be selected from, but are not limited to, the group consisting of RNAs, small RNAs, functional RNAs, transfer RNAs or their fragments, circular RNAs, large RNAs, coding RNAs, non-coding RNAs, peptides, proteins, DNA sequences, metabolites, lipids, carbohydrates, internal controls, cell counts, tissue weights, equivalent sample measures, and any combination or subset thereof.

[0166] In some embodiments, a “treatment” or “intervention” refers to any preventive and / or therapeutic medical action to prevent, treat, postpone, or mitigate a disease or disorder, and / or the symptoms related to a disease or disorder.

[0167] In some embodiments, the “mathematical method” includes, but is not limited to all algorithms implemented in the caret R package (v.6.0-94, [2]), e.g. linear, logistic, generalized linear (mixed), and Cox regression, matrix factorization, independent Principal Component Analysis (PCA), Partial Least Squares (or Projection to Latent Space - PLS), and Partial Least Squares Discriminant Analysis (PLS-DA).

[0168] In some embodiments, the “logical method” includes, but is not limited to decision trees or random forests.

[0169] In some embodiments, the “machine learning” methods include, but are not limited to neuronal networks, deep learning, elastic net, lasso regression, support vector machines (SVN), genetic algorithms, c5.0-method, etc.

[0170] Analysis of the data and the results derived therefrom are typically performed using computer hardware operating according to defined algorithms and programs. Therefore, certain embodiments employ processes involving data stored in or transferred through one or more computer systems or other processing systems. Embodiments of the invention also relate to an apparatus for performing these operations. This apparatus may be specially constructed for the required purposes, or it may be a general-purpose computer (or a group of computers) selectively activated or reconfigured by a computer program and / or data structure stored in the computer. In some embodiments, a group of processors performs some or all of the recited analytical operations collaboratively (e.g., via a network or cloud computing) and / or in parallel. A processor or group of processors for performing the methods described herein may be of various types including microcontrollers and microprocessors such as programmable devices (e.g., CPLDs and FPGAs) and other devices such as gate array ASICs, digital signal processors, and / or general purpose microprocessors.

[0171] In addition, certain embodiments relate to tangible and / or non-transitory computer readable media or computer program products that include program instructions and / or data (including data structures) for performing various computer-implemented operations. Examples of computer-readable media include, but are not limited to, semiconductor memory devices, magnetic media such as disk drives, magnetic tape, optical media such as CDs, magneto-optical media, and hardware devices that are specially configured to store and perform program instructions, such as read-only memory devices (ROM) and random access memory (RAM). The computer readable media may be directly controlled by an end user or the media may be indirectly controlled by the end user. Examples of directly controlled media include the media located at a user facility and / or media that are not shared with other entities. Examples of indirectly controlled media include media that is indirectly accessible to the user via an external network and / or via a service providing shared resources such as the "cloud." Examples of program instructions include both machine code, such as produced by a compiler, and files containing higher level code that may be executed by the computer using an interpreter.

[0172] According to one embodiment, the use of a computer-implemented method is also comprised herein.

[0173] In some embodiments, the analysis of the normalized levels can include feature selection and one or more internal cross-validations.

[0174] In some embodiments, the “biological data” includes, but is not limited to, species, strain, sub-strain, genetic modification, knock-out strain, height, weight, BMI, sex, ethnic origin. In some embodiments, the “clinical data” includes, but is not limited to, pathological specifications, histology, stage, grade, anamnesis, blood picture, blood pressure, vital signs, medical background, clinical status, pedigree anamnesis.

[0175] In some embodiments, the “environmental data” includes, but is not limited to, temperature, nutrition, air pressure, air composition.

[0176] In some embodiments, the “biographical data” includes, but is not limited to, age, gender, origin, ethnic origin, genealogical tree, marital status, smoker status, alcohol abuse status, illegal drug abuse status, drug status.

[0177] According to one embodiment, the multiple data sets and / or states may be selected and “reduced in dimensions” to transform the data from a high-dimensional space into a low-dimensional space to facilitate feature selection using principal component analysis (PCA), kernel PCA, graph-based kernel PCA, linear discriminant analysis (LDA), generalized discriminant analysis (GDA), t-SNE, Uniform manifold approximation and projection (UMAP), and matrix factorization (NMF), random forests, or gradient boosting machines, auto-encoders, or similar methods, and / or any combination of these. Suppression of batch effects on the feature selection procedure is achieved by applying standard controlling procedures involving computing of correlation metrics, regression analysis, (hierarchical) clustering of features based on similarity and testing resulting clusters against known possible confounding variables like sequencing batch, sequencing technology, depth of coverage, age of sample, sample sex (wherever applicable) an any other example.

[0178] As used herein, the term “swap” refers to the procedure of collecting any sample that has been obtained by contacting a surface, liquid, or tissue with a sterile and clean cotton bud, or comparable products.

[0179] As used herein, the term “smear” refers to the procedure of collecting of any sample by contacting a surface, liquid, or tissue with a sterile and clean small brush, spatula, cotton bud, or comparable products.

[0180] As used herein, the term “cell cluster” refers to spheroids, or organoids, wherein the former refers to organ-specific cells and the latter refers to broad-ranging cells such as e.g. from tumor tissue, nervous tissue or hepatocytes.

[0181] As used herein, the term “aspirate” refers to any sample that has been obtained by drawing a liquid by suction. As used herein, the term “cell free supernatant” refers to a supernatant that is substantially free of cells. Specifically, “substantially” means “completely”.

[0182] As used herein, the term “cellular sternness” refers to developmental stage of a stem cell defined on the ability for self-renewal and the ability for multilineage differentiation.

[0183] Small RNA “libraries” can be generated using library preparation kits well known in the art, such as the NEBNext® Small RNA Library Prep Set (NEB, USA) or the CleanTag SmalIRNA library preparation kit (TriLink, USA). Usually, RNA is first reverse transcribed into DNA followed by PCR amplification. PCR amplification can be performed using barcoded primers, such as the barcoded primers from Illumina for small RNA sequencing. Before sequencing, PCR products are purified and size selected using protocols well known in the art. For example, PCR products can be purified using the QiaQuick protocol from Qiagen, size selected by magnetic beads, electrophoresis or column-based methods and can subsequently be size checked by suitable methods, such as e.g. capillary electrophoresis (QSEP from BiOptic inc. or Agilent 4200 TapeStation System) or Agilent Bioanalyzer chips. In a preferred embodiment, the small RNA libraries are mirxon libraries.

[0184] Specifically, next-generation sequencing can be performed on any suitable platform, such as e.g. the Illumina NextSeq 500. Sequencing reads are usually adaptertrimmed, quality checked and edited according to bioinformatics methods well known in the art to prepare for further use.

[0185] According to one embodiment, the library described herein comprises mirxons, specifically 10 or more mirxons, specifically the library comprises about 10-6000 mirxons, specifically 100-6000, 200-6000, 300-6000, 400-6000, 500-6000, 600-6000, 700-6000, 800-6000, 900-6000, 1000-6000, 1100-6000, 1200-6000, 1300-6000, 1400- 6000, 1500-6000, 1600-6000, 1700-6000, 1800-6000, 1900-6000, 2000-6000, 2100-

[0186] 6000, 2200-6000, 2300-6000, 2400-6000, 2500-6000, 2600-6000, 2700-6000, 2800-

[0187] 6000, 2900-6000, 3000-6000, 3100-6000, 3200-6000, 3300-6000, 3400-6000, 3500-

[0188] 6000, 3600-6000, 3700-6000, 3800-6000, 3900-6000, 4000-6000, 4100-6000, 4200-

[0189] 6000, 4300-6000, 4400-6000, 4500-6000, 4600-6000, 4700-6000, 4800-6000, 4900- 6000, 5000-6000, 5100-6000, 5200-6000, 5300-6000, 5400-6000, 5500-6000, 5600- 6000, 5700-6000, 5800-6000, or 5900-6000 mirxons, more specifically 10, 11 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 , 32, 33, 34, 35, 36, 37, 38, 39, 40, 41 , 42, 43, 44, 45, 46, 47, 48, 49, 50, 51 , 52, 53, 54, 55, 56, 57, 58, 59, 60, 61 , 62, 63, 64, 65, 66, 67, 68, 69, 70, 71 , 72, 73, 74, 75, 76, 77, 78, 79, 80, 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91 , 92, 93, 94, 95, 96, 97, 98, 99, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220, 230, 240, 250, 260, 270, 280, 290, 300, 310,

[0190] 320, 330, 340, 350, 360, 370, 380, 390, 400, 410, 420, 430, 440, 450, 460, 470, 480,

[0191] 490, 500, 510, 520, 530, 540, 550, 560, 570, 580, 590, 600, 610, 620, 630, 640, 650,

[0192] 660, 670, 680, 690, 700, 710, 720, 730, 740, 750, 760, 770, 780, 790, 800, 810, 820,

[0193] 830, 840, 850, 860, 870, 880, 890, 900, 910, 920, 930, 940, 950, 960, 970, 980, 990,

[0194] 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300,

[0195] 2400, 2500, 2600, 2700, 2800, 2900, 3000, 3100, 3200, 3300, 3400, 3500, 3600, 3700,

[0196] 3800, 3900, 4000, 4100, 4200, 4300, 4400, 4500, 4600, 4700, 4800, 4900, 5000, 5100,

[0197] 5200, 5300, 5400, 5500, 5600, 5700, 5800, 5900 or 6000 mirxons.

[0198] According to an alternative embodiment, the library contains data sets of mirxons, such as the data sets from their expression levels. Specifically the library comprises about 10-6000 mirxons, specifically 100-6000, 200-6000, 300-6000, 400-6000, 500- 6000, 600-6000, 700-6000, 800-6000, 900-6000, 1000-6000, 1100-6000, 1200-6000, 1300-6000, 1400-6000, 1500-6000, 1600-6000, 1700-6000, 1800-6000, 1900-6000,

[0199] 2000-6000, 2100-6000, 2200-6000, 2300-6000, 2400-6000, 2500-6000, 2600-6000,

[0200] 2700-6000, 2800-6000, 2900-6000, 3000-6000, 3100-6000, 3200-6000, 3300-6000,

[0201] 3400-6000, 3500-6000, 3600-6000, 3700-6000, 3800-6000, 3900-6000, 4000-6000,

[0202] 4100-6000, 4200-6000, 4300-6000, 4400-6000, 4500-6000, 4600-6000, 4700-6000,

[0203] 4800-6000, 4900-6000, 5000-6000, 5100-6000, 5200-6000, 5300-6000, 5400-6000,

[0204] 5500-6000, 5600-6000, 5700-6000, 5800-6000, or 5900-6000 mirxons, more specifically 10, 11 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29, 30, 31 , 32,

[0205] 33, 34, 35, 36, 37, 38, 39, 40, 41 , 42, 43, 44, 45, 46, 47, 48, 49, 50, 51 , 52, 53, 54, 55,

[0206] 56, 57, 58, 59, 60, 61 , 62, 63, 64, 65, 66, 67, 68, 69, 70, 71 , 72, 73, 74, 75, 76, 77, 78,

[0207] 79, 80, 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91 , 92, 93, 94, 95, 96, 97, 98, 99, 100, 110,

[0208] 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220, 230, 240, 250, 260, 270, 280,

[0209] 290, 300, 310, 320, 330, 340, 350, 360, 370, 380, 390, 400, 410, 420, 430, 440, 450,

[0210] 460, 470, 480, 490, 500, 510, 520, 530, 540, 550, 560, 570, 580, 590, 600, 610, 620,

[0211] 630, 640, 650, 660, 670, 680, 690, 700, 710, 720, 730, 740, 750, 760, 770, 780, 790,

[0212] 800, 810, 820, 830, 840, 850, 860, 870, 880, 890, 900, 910, 920, 930, 940, 950, 960,

[0213] 970, 980, 990, , 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000, 2100, 2200, 2300, 2400, 2500, 2600, 2700, 2800, 2900, 3000, 3100, 3200, 3300, 3400, 3500, 3600, 3700, 3800, 3900, 4000, 4100, 4200, 4300, 4400, 4500, 4600, 4700, 4800, 4900, 5000, 5100, 5200, 5300, 5400, 5500, 5600, 5700, 5800, 5900 or 6000 data sets.

[0214] The present invention also provides a composition containing mirxons. According to a specific embodiment, the backbones of said mirxons are replaced by molecules not naturally occurring in RNA molecules, specifically to increase stability and robustness of the mirxons. Alternatively, RNA stabilizers may be added to stabilize the integrity of the mirxons, specifically during transport and storage at ambient temperatures. Stabilizers are well known in the art, such as lipid-based nanoparticles, polyethylene glycol, polysorbate 80, RNAIater™, RNAstable® from Merck, RNA rescue™ stabilizing reagents from Genetech Biotech.

[0215] A pharmaceutical preparation comprising the mirxons is also provided herein, specifically present in a pharmaceutically effective amount. The pharmaceutical preparation described herein can be administered or mixed with any pharmaceutically acceptable substances or carriers or excipients as known in the art. These can be for example, but are not restricted to water; parenteral solutions; neutralizing agents like NaOH, KOH; stabilizers; buffers, e.g. acetate buffer, phosphate buffer, citrate buffer; solvents; DMSO; or saline.

[0216] The term “therapeutically effective amount”, of the mirxons of the present invention is a quantity sufficient to, when administered to a subject, effect beneficial or desired results, including clinical results, and, as such, an effective amount or synonym thereof depends upon the context in which it is being applied.

[0217] The pharmaceutical preparation can be administered to the subject by any mode applicable and known in the art. More specifically, enteral, intravenous, intraperitoneal, intramuscular, or subcutaneous administration or by spray, salve, emulsion, is provided.

[0218] Because mirxons are involved in a huge variety of diseases or disorders, the composition described herein may also be applicable for the treatment of said disorders. Said malignancies may be, but are not limited to malignant disease, cardiovascular disease, endocrine disorder and disease, immunological disease, mental disorder, psychological disorder and disease, infection, malnutrition, intoxication, or injury.

[0219] The present invention also encompasses the following embodiments: 1 . Method for in vitro determining a change of a biological state of a subject, comprising the steps

[0220] (a) providing at least two mirxons from a reference sample, which represent a set of biological states, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), wherein

[0221] N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0222] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches,

[0223] (b) providing the same mirxons of (a) from a sample of the subject,

[0224] (c) quantifying the levels of one or more of said mirxons and correlating the levels of the mirxons of (a) with the levels of the mirxons of (b), wherein any deviation of said levels reflects differences in said biological states, which are indicative for the change of the biological state to be determined, and

[0225] (d) normalizing the levels of the mirxons of (c) to the levels of one or more normalizers from the samples, or to a subset of mirxons or all mirxons of (a) or to other RNAs, wherein the levels of said normalizers are most constant over the biological states.

[0226] 2. The method of embodiment 1 , wherein the reference sample is from one or more subjects of the same species, specifically said sample is from healthy subjects, subjects of one or more biological states.

[0227] 3. Method for in vitro monitoring the treatment of a subject, comprising the steps

[0228] (a) providing at least two mirxons from a reference sample, which represent a set of biological states, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), wherein

[0229] N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0230] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches,

[0231] (b) providing the same mirxons of (a) from a sample of the subject, (c) quantifying the levels of one or more of said mirxons and correlating the levels of the mirxons of (a) with the levels of the mirxons of (b), wherein any deviation of said levels reflects differences in said biological states, which are indicative for the change of the biological state to be determined, and

[0232] (d) normalizing the levels of the mirxons of (c) to the levels of one or more normalizers from the samples, or to a subset of mirxons or all mirxons of (a) or to other RNAs, wherein the levels of said normalizers are most constant over the biological states.

[0233] 4. The method of any one of embodiments 1 to 3, wherein the normalized levels are analyzed with a mathematical or logical method to provide information on the probability of a biological state, condition, or change of the subject.

[0234] 5. The method of embodiment 4, wherein the normalized levels are combined with further RNA levels, or biological, clinical, environmental, experimental, or biographical data.

[0235] 6. The method of any one of embodiments 1 to 5, wherein the normalizers are selected from the group consisting of RNAs, small RNAs, functional RNAs, transfer RNAs or their fragments, circular RNAs, large RNAs, coding RNAs, non-coding RNAs, peptides, proteins, DNA sequences, metabolites, lipids, carbohydrates, internal controls, cell counts, tissue weights, and any combination or subset thereof.

[0236] 7. The method of any one of embodiments 1 to 6, wherein the difference in the levels of the mirxons is determined by endpoint, quantitative or digital PCR, miRCURY assays, sequencing, microarrays, Luminex nucleic acid assays, Nanostring assays, or other hybridization-based techniques.

[0237] 8. The method of any one of embodiments 1 to 7, wherein the sample is selected from the group of blood, extracellular vesicle, tissue, cell, cell cluster, organoid, spheroid, cell aggregate, ascites, pleural effusion, stool, urine, sperm, saliva, menstrual blood, bile, bone marrow aspirate, breast milk, cerebrospinal fluid, plasma, serum, sweat, sputum, synovial fluid, lymph, oral swab, nasal swap, oropharyngeal swap, nasopharyngeal swap, vaginal swab, urethral swab, cervical swab or lavage, smear of a surface, and body cavity or cell free supernatant of in vitro grown tissues, cells or cell aggregates. 9. Use of the method of any one of embodiments 1 to 8, for monitoring a subject, specifically for monitoring the development, progression, or change of a biological state.

[0238] 10. The method of any one of embodiments 1 to 9, wherein the biological states are selected from the group consisting of biological, physical, chemical, clinical, developmental and pathological states, specifically malignant disease, cardiovascular disease, endocrine disorder and disease, immunological disease, mental disorder, psychological disorder and disease, infection, malnutrition, intoxication, injury, cellular sternness, cellular developmental state, cellular degree of differentiation, or a selection thereof.

[0239] 11. The method of any one of embodiments 1 to 10, wherein the mirxons comprise mismatches at the C nucleotides of the (CGG)e (SEQ ID NO: 3) nucleotide sequence.

[0240] 12. The method of any one of embodiments 1 to 10, wherein the mirxons comprise mismatches at the G nucleotides of the (CCG)e (SEQ ID NO: 4) nucleotide sequence.

[0241] 13. The method of any one of embodiments 1 to 12, wherein the mirxons comprise mismatches which are one-base deletions or insertions.

[0242] 14. A library comprising two or more mirxons, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx(SEQ ID NO: 2), wherein

[0243] N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0244] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

[0245] 15. The library of embodiment 14, wherein said library is constructed by isolating and selecting one or more mirxons from one or more samples.

[0246] 16. A reference sample library comprising data sets of at least two mirxon levels, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)eNy(SEQ ID NO: 1) or Ny(CCG)6Nx(SEQ ID NO: 2), wherein

[0247] N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0248] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

[0249] 17. Composition comprising one or more mirxons, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)eNy (SEQ ID NO: 1) or Ny(CCG)6Nx(SEQ ID NO: 2), wherein

[0250] N can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and

[0251] (CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

[0252] 18. The composition of embodiment 17, wherein the bases or backbone of said mirxons are replaced by molecules not naturally occurring in RNA molecules, modified ribose sugars, or amide linkages.

[0253] 19. The composition of embodiment 18, wherein said molecules are deoxyribonucleotide bases, artificial bases, locked nucleic acid (LNA) bases, or equivalent molecules.

[0254] 20. Pharmaceutical composition comprising the composition of any one of embodiments 17 to 19 and pharmaceutical acceptable excipients.

[0255] 21. The composition of any one of embodiments 17 to 19 or the pharmaceutical composition of embodiment 20 for use in the treatment of a disease or disorder, specifically selected from the group consisting of malignant disease, cardiovascular disease, endocrine disorder and disease, immunological disease, mental disorder, psychological disorder and disease, infection, malnutrition, intoxication, injury.

[0256] The examples described herein are illustrative of the present invention and are not intended to be limitations thereon. Many modifications and variations may be made to the techniques described and illustrated herein without departing from scope of the invention. EXAMPLES

[0257] Example 1

[0258] Methods

[0259] With small RNA sequencing 43 primary ovarian cancer tissues [3] were analyzed with miRDeep* v3.8, which included the prediction of new miRNAs (not known from miRBase) using the combined reads larger than 16 nucleotides from all samples, totaling to more than half a billion reads [4], MirDeep* predicted 4,409 new miRNAs and confirmed 768 known from miRBase v20 (https: / / www.mirbase.org / ). These data will be called “OVARIAN” from now on.

[0260] 82 small RNA sequencing data were downloaded from 36 different human tissues from the ENCODE consortium [5] and combined all complete reads (i.e. surrounded by the library adapter sequences) between 17 to 24 nucleotides (totaling to 1.1 x 109reads) and predicted miRNAs using the miRPipe pipeline for the HG38 human genome (which uses also miRDeep* for prediction) [6]. 15,024 new putative miRNAs were predicted and 1 ,350 miRBase v22.1 ones confirmed. This set will be called “ENCODE” from now on.

[0261] As described by Bachmayr-Heyda et al. [4] new miRNAs were predicted from small RNA-sequencing reads of 43 EpCAM enriched primary ovarian cancer cells using miRDeep* v3.8 (OVARIAN dataset). MiRNAs consisting of (CGG) or (CCG)-repeats were very abundant and were systematically analyzed. All mature miRNA sequences were analyzed for the sequence (CGG)e (SEQ ID NO: 3) and the reverse complement (CCG)e (SEQ ID NO: 4) with up to six mismatches using function matchPattern from R- package Biostrings (v2.68.1) and combined in eight classes — (CGG)e / (CCG)6 ((SEQ ID NO: 3) / (SEQ ID NO: 4)) with zero to six mismatches and without (CGG)e / (CCG)6 ((SEQ ID NO: 3) / (SEQ ID NO: 4)) sequence at all. A similar analysis was done with trimmed small RNA-sequencing reads from 82 samples of 36 different human tissues from the ENCODE consortium [5], [7] using the miRPipe pipeline for the HG38 human genome (which uses also miRDeep* for prediction; ENCODE dataset) [6]. All predicted miRNAs were analyzed for other possible trinucleotide repeats up to six mismatches (shown are only up to five mismatches), reporting only one of the two possible directions ((TTT)e (SEQ ID NO: 5) means (TTT)6and (AAA)6(SEQ ID NO: 6)).

[0262] The origin of the predicted miRNAs, respectively of the corresponding predicted pre-miRNA stem-loops, was determined with respect to gene annotations of the human genome HG38 and every miRNA and corresponding pre-miRNA was classified as either originating from first exon / promoter, intronic, or intergenic using R-package annotatr (v 1.26.0). The relative ratio origin from first exon / promotor of each (CGG)e (SEQ ID NO: 3) mismatch class was calculated and plotted. Most of (CGG)e (SEQ ID NO: 3) miRNAs originated from first exons, therefore, and analogous to the term mirtrons, this class of miRNAs was named mirxons. Histograms of the lengths of all first exons containing a mirxon and not containing a mirxon were plotted and compared. Genes containing a mirxon and genes containing mirtrons (miRNAs originating from introns of other genes) were functionally annotated and enriched categories presented as p-value color-coded word cloud using R-package GOsummaries (v2.36.0).

[0263] To assess the expression of all predicted ENCODE miRNAs in all datasets analyzed here the unitas pipeline (v1.8.0) [8] was employed with two changes. 1) After downloading of all needed databases from the Unitas program the miRBase_mature and the miRBase_hairpin_Human.fas files were updated with the mature and pre-miRNA sequences of the predicted ENCODE miRNAs not already present in the miRBase 22.1 . 2) The run was performed without filtering low-complexity sequences (~skip_dust). Parameter “-trim” <known adapter sequencer or “-trim ?”, if adapter sequence is not known, was used as needed or omitted if reads were already trimmed. Subsequently, the Unitas output file miR-table_Human.simplified.txt was used. The variance of the expression of mirxons, miRBase known, and other predicted miRNAs was calculated by the mean absolute deviation (MAD) per median and plotted against the median expression.

[0264] To validate the expression of two specific mirxon types, ending either CGGCGGCGGCGGCGGA (SEQ ID NO: 7) or CGGCGGCGGCGGCGGGA (SEQ ID NO: 8) miRCURY LNA miRNA PCR Assays using locked nucleic acids (LNA) were developed together with QIAGEN (Hilden, Germany) and can be purchased via GeneGlobe (Cat) Numbers YCP0054165 or YCP0054166, respectively (Figure 1). Figure 1 shows the primer sequences and melting temperatures for the QIAGEN LNA based RT-qPCR systems for two different mirxons, differentiating only in the next to last nucleotide (i.e. GGA and GGGA). Figure 1 further refers to the following sequences: GCCTTTTTTTTTTTTTTT u-TAG (SEQ ID NO: 9) CGGCGGCGGCGGCGGAAAAAAAAAAAAAAAAAAAAAAA (SEQ ID NO: 10) GGCGGCGGCGGCGGA (SEQ ID NO: 11) GCCTTTTTTTTTTTTTTT u-TAG (SEQ ID NO: 12) CGGCGGCGGCGGCGGGAAAAAAAAAAAAAAAAAAAAAAA (SEQ ID NO: 13) GCGGCGGCGGCGGG (SEQ ID NO: 14)

[0265] PCR-conditions were optimized and are as follows: [ 95 °C 15 sec; 70 °C 30 sec ] 40x.

[0266] (CGG)e (SEQ ID NO: 3) mirxons were silenced by the QIAGEN miRCURY LNA miRNA Power Inhibitor technology using the LNA-based inhibitor (CCG)e (SEQ ID NO: 4) and the miRCURY LNA miRNA Power Inhibitor Negative Control A as control. Uptake of the power inhibitors was accomplished by gymnosis without added transfection reagent in complete medium on cells with about 25 % confluency in six-well plates and cells were further grown for 72 h. The following cell lines were used in triplicates for each condition; ES2, TyKNU, OVCAR3, and DLD and said cell lines were grown in RPMI / 10%FCS, aMEM / 10%FCS, RPMI / 20%FCS, and DMEM / 10%FCS medium, respectively. Cells were harvested directly with QIAzol Lysis Reagent (QIAGEN) from the plates and small and large RNAs were isolated with the miRNeasy Mini Kit (QIAGEN, two fraction protocol). Libraries for RNA sequencing were prepared with the NEBNext® Ultra™ II Directional RNA Library Prep Kit (New England Biolabs (NEB), Ipswich, Massachusetts, U.S.A.) and the NEBNext® rRNA Depletion Kit (NEB) and sequenced on an Illumina HighSeq 4000 sequencer for 2 x 150 bases paired ends. Obtained reads were mapped to the HG38 genome using STAR-2.7.10a and reads counted into the Gencode gene model v40 by the featureCounts program from the subread package (v2.0.6). Counted reads were normalized by voomWithQualityWeights function from R- package limma (v 3.56.2) using the “cyclicloess” method. Differentially expressed genes between the (CCG)e (SEQ ID NO: 4) silenced and the Negative Control A treated cells were determined over all four cell lines, including cell line as confounding factor in the model.

[0267] Using this differentially gene expression result a gene-signature comprised of 565 genes, 297 higher-expressed in the (CGG)e (SEQ ID NO: 3) silenced samples and 268 higher-expressed in the control samples, was extracted using R-package CellMix (v1 .6.2) with the Abbas-method and median expression of the 297 up-regulated genes minus median expression of the 268 down-regulated genes calculated for all gene expression data sets used in this work. This 565-gene-signature predictor was used for multiple Cox regression analyses for overall survival in an own gene expression dataset (GSE49997 [3], together with several clinicopathologic confounders) and as meta- analysis in 16 publicly available ovarian cancer gene expression datasets (as implemented in R-package curatedOvarianData v1.38.0 using all with survival data available) (with Federation of Gynecology and Obstetrics (FIGO)) stage and age as confounding parameters).

[0268] Results

[0269] Discovery

[0270] Figure 2 shows the sequence logos of several mirxon candidates. Figure 2A shows a sequence with four mismatches. Figure 2B shows a model sequence with its stem-loop pre-miRNA. The sequence thereof is as follows: GGGGCGGCGGCGGCGGCGGAGGCGGCCGGGTCCTGCCGCGGGCAGCGCGCG CTG (SEQ ID NO: 20). Figure 2C shows many other candidate sequences, built from all of their isomiRs as found in the OVARIAN small RNA sequencing data. Looking on all the sequences of the new miRNAs one sequence was vastly over-represented, namely, the trinucleotide repeat (CGG) concatenated up to six times ((CGG)e (SEQ ID NO: 3); containing up to six mismatches) (Figure 2). Specifically, from all putative miRNAs 109, 139, 111 , 42, 21 , 19 (one known in miRBase), and 133 (five from miRBase) consisted of the (CGG)e-repeat (SEQ ID NO: 3) with zero to six mismatches, respectively. Hundreds of (CGG)e (SEQ ID NO: 3) containing miRs were predicted, partly overlapping with the ones from the OVARIAN data (Figure 3). Figure 3 shows the overlap of newly predicted mirxons up to five mismatches (left panel) and with exactly six mismatches (right panel) between the OVARIAN and the ENCODE small RNA-sequencing dataset. Figure 3 highlights that in total 1399 putative mirxons were predicted. To systematically determine the composition of these putative miRNAs concerning trinucleotide repeats, all predicted mature miRNA sequences were analyzed for all 64 possible trinucleotides (AAA to TTT) concatenated six times (NNN)e (SEQ ID NO: 28) with up to six mismatches. Results for reverse complements (e.g. (CGG) (CCG)) were identical and are therefore not shown anymore. Figure 4A shows the numbers of miRNAs (as predicted from two large small RNA-sequencing data sets, OVARIAN and ENCODE) whose sequences are comprised of trinucleotide repeats (six times repeated, (NNN)e (SEQ ID NO: 28) and up to five mismatches). The three trinucleotides which represent all the type (CGG) (i.e. (CGG) / GCG / GGC) and their reverse complements (CCG / CGC / GCC), which were counted together with their reverse complements, were the most abundant ones. As expected, all three rotations representing the (CGG)-repeat, namely (CGG), (GCG), and, (GCG) (the corresponding reverse complement sequences, not shown) were by far the most abundant sequences among all possible trinucleotides. Therefore, this class of putative miRNAs (CGG)e (SEQ ID NO: 3)-containing up to six mismatches will be called mirxons or (CGG)e-miRNAs (SEQ ID NO: 3). Only considering five mismatches, the trinucleotides GGG, TGG, ACC, and GTG showed numbers over 100. Most of these can be explained by the known hot spot mutation of methylated CGs to TGs, which are very abundant in the human genome. Figure 5 shows the sequence logos of putative mirxons up to four mismatches (If sequences are used with more than four mismatches the used method for multiple sequence alignment cannot converge the sequence alignment to only one overlapping sequence). The sequence logo for two mismatches is reverse complement as in this class the majority of mirxons are reverse complement. Figure 5 indicates the CG to TG mutation as the most frequent difference as inferred from the known hotspot mutation of (methylated) CGs in the human genome.

[0271] Characterization of (CGG) s (SEQ ID NO: 3) small RNAs

[0272] Most of the known miRNA genes coding for the pri-miRNAs are intragenic (in respect to protein-coding or non-miRNA non-coding genes), some of them are intronic, and called therefore mirtrons. Annotating the (CGG)e (SEQ ID NO: 3) containing putative mature miRNAs and their precursor sequences (their pre-miRNA stem-loops) revealed a very high concentration of these in first exons or direct vicinity in front of first exons (promotor regions) of mostly protein-coding genes. Figure 4B shows the percentages of miRNAs (left) and their predicted pre-miRNAs stem-loops (right) which are located in first exons or promotor regions of (protein-coding) genes. The Y-axis of Figure 4B shows the percentage. The X-axis shows mirxons, comprised of the (CGG)e (SEQ ID NO: 3) sequence up to six mismatches (MM0-MM6) and other miRNAs not comprised of the (CGG)e (SEQ ID NO: 3) sequence (None and sNone). In other words, Figure 4B shows the percentages of mature (CGG)e (SEQ ID NO: 3) miRNAs (mMM) and their stem-loops (sMM) up to six mismatches (xMM0..6) and all other miRNAs (xNone) in first exons or directly upstream. Up to four mismatches more than 80 % of the miRNAs are located in first exons or promotor regions, compared to ~20 % of all other miRNAs, including thousands of newly predicted ones. Based on the term mirtrons for miRNAs residing in introns of other genes (the term is composed of the first part of m / 'RNA and the last part of ntron), the putative (CGG)e (SEQ ID NO: 3) miRNA class will be called mirxons, combining the terms m / 'RNA and exon. In accordance with this, first exons which contain such a mirxon are substantially larger compared to other first exons of (protein-coding) genes, i.e. median 419 bp versus median 154 bp, respectively (Figure 4C). Figure 4C shows the density distributions and median values of the length of first exons containing mirxons (right peak (red) with right number below the x-axis), i.e. 469 bp, and all other first exons not containing mirxons (left peak (blue) with left number below the x-axis), i.e. 154 bp.

[0273] Figure 4D shows the enriched biological annotations for mirxon containing (protein-coding) genes and mirtron containing (protein-coding) genes. Figure 4D shows Word clouds of Gene Ontology (GO) terms and KEGG (https: / / www.genome.jp / kegg / ) pathways over-represented in genes with mirxons (top) and genes with mirtrons (miRNAs located in introns of other genes) (bottom). MiRNAs with own (pri-miRNA) genes cannot be analyzed this way, as there are no protein-coding genes, thus functions associated. The former include terms related to transcription regulation and tissue morphogenesis and development and the latter cytoskeletal proteins, cell morphogenesis and neuron development.

[0274] Expression

[0275] To assess the expression and the variance of the expression of the newly predicted and miRBase-known miRNAs, including the newly discovered mirxons, the unitas pipeline [8] was employed, with two modifications: Firstly, the tables of known human mature miRNAs and their stem-loop sequences (downloaded from the miRbase v22.1 [9]) were supplemented with the newly predicted mature and stem-loop- sequences and secondly, the run was done without filtering out low complexity sequences (syntax: “-skip_dust”). After loading, filtering of reliably expressed reads (i.e. reads per million > 0.5 in more than halve of samples), normalizing by the total number of miRNA reads and taking the binary logarithm with an offset of 0.5, the variation of the expression of every miRNA over the human tissues was calculated by the MAD per median expression and grouped as follows: mirxons up to six mismatches, miRBase miRNAs and newly predicted putative miRNAs excluding the mirxons. Figure 6A shows the variability of mirxon (up to six mismatches) expressions (top line (red) with confidence intervals in the main plot) compared to the expression of miRBase known miRNAs (bottom line (green) with confidence intervals in the main plot) and other newly predicted miRNAs (middle line (blue) with confidence intervals in the main blot) as inferred from small RNA-sequencing data of different human tissues. X-axis shows the median expression of the miRNA and the Y-axis the median absolute deviation (MAD) per median value over all tissues. In Figure 6A the MAD / median values are plotted against the median expressions. Most miRBase miRNAs showed a low MAD / median variation (below one) by rather high median expression values with only a few miRNAs with higher variation values. Whereas approximately half of the newly predicted putative miRNAs showed low MAD / median values, comparable to the ones from the miRBase, the other half showed nearly twice as high MAD / median values, indicating a high tissue specificity or at least expression variation. In total, the newly predicted miRNAs showed lower median expression values. Interestingly, the mirxons showed mostly high MAD / median values at rather low median expression values. Mirxons seem to be expressed rather low and more tissue-specific than miRBase miRNAs. To assess the relative expression of mirxons in the different human tissues and exosomes, all reads belonging to mirxons (up to six mismatches) were divided by all reads belonging to any miRNA and the binary logarithm was taken. Figure 6B shows the results sorted according expression ratio in the corresponding tissue. The Iog2 ratio of reads associated to mirxons compared to reads from all miRNAs in different human tissues are shown. The difference of the highest ratio to the lowest one is approx. 64 times in linear space (i.e. Iog2 6, from -11 to -5, the value -4 is from extracellular vesicles from senescent human umbilical cord-derived mesenchymal stem cells). The Y-axis of Figure 6B shows the Log2 ratio mirxons to all miRNAs. Figure 6C and D show the same analyses of an independent small RNA-sequencing dataset of different human cell types. The Y-axis of Figure 6D shows the Log2 ratio mirxons to all miRNAs. Again, mirxons are the most variable expressed ones (top (red) line with confidence intervals in the main plot of Figure 6C compared to the middle (green) line of miRBase miRNAs and the bottom (blue) line of other small RNAs) with approx. 64 times difference between the cell type with the lowest ratio and the one with the highest ratio. The variability was high, from a Iog2-ratio of -11 (i.e. metanephros) to -5 (i.e. subcutaneous adipose tissue) and -4 (i.e. exosomes from senescent human umbilical cord mesenchymal stem cells), which means linearly about 0.05 % to 3 % (6.3 % for exosomes) of miRNA reads belong to mirxons, leading to a wide range of about 60 times difference between human tissue with the lowest and highest ratio. A clear interpretation concerning a specific common characteristic of the tissues with low or high ratios, e.g. low or high proliferating or differentiated from a specific germ layer or progenitor (stem) cell, is not obvious. A similar analysis using 38 human cell types revealed a very similar result. The range was again from Iog2 -4 (colon epithelial cells) to -11 (renal epithelial cells) and again higher MAD / median values for mirxons at lower median expressions compared to miRBase and other new miRNAs.

[0276] Validation of the expression of newly predicted putative mature miRNAs and mirxons

[0277] Bachmayr-Heyda et al. [4] validated the expression of 22 newly predicted OVARIAN miRNAs using LNA oligo based RT-qPCR systems (QIAGEN former Exiqon), with a success rate of 81.8 % (OVARIAN miR-1003-3p / ENCODE miR-7056-3p (2 mismatches): GCGGCGGCGGCGGCGGUG (SEQ ID NO: 15) from gene UBQLN1 / 1stexon, OVARIAN miR-2769-5p / ENCODE miR-13144-5p (1 mismatch):

[0278] GCGGCGGCGGCGGCGGCGA (SEQ ID NO: 16) from gene CHD1 / 1stexon, and OVARIAN miR-3475-3p / ENCODE miR-9032-3p (3 mismatches):

[0279] UGGGGGCGGCGGCGGGGG (SEQ ID NO: 17) from gene IGSF9B / 1stintron). Figure 7 shows the genome plots of the genomic loci of the three mirxons which were experimentally verified by RT-qPCR in [4], All three are either in a first exon of a protein coding gene (UBQLN1 and CHD1) or short downstream of a first exon in the intron (IGSF9B). Figure 7 shows the genomic loci of the 1stexons with its neighborhoods for these genes with the positions of the mirxons and repetitive (CGG)e (SEQ ID NO: 3) sequences. Two LNA-based systems were constructed for two very similar OVARIAN mirxons, i.e. (CGG)eA (SEQ ID NO: 18) (“GGA”) present in 20 mirxon loci whose mature miRNAs end exact with this sequence and (CGG)eGA (SEQ ID NO: 19) (“GGGA”) present in 19 mirxon loci, as above (Figure 1). Both were relatively high expressed in the ovarian tumor tissues and a genome-wide search for exactly these sequences revealed 112 (CGG)eA (SEQ ID NO: 18) sequences (52 forward and 60 reverse in the human genome) and 79 (CGG)eGA (SEQ ID NO: 19) sequences (44 forward and 35 reverse). Nearly all of them are located in first exons or promoter regions of (proteincoding) genes with significantly longer first exons, as described above. Both systems worked and yielded RT-qPCR results from cell line, fresh tumor tissue, and FFPE tissue derived RNAs.

[0280] Silencing of a mirxon and impact on gene expression

[0281] Using the QIAGEN miRCURY LNA miRNA Power Inhibitor technology a Locked Nucleic Acid (LNA) based inhibitor was designed raised against the (CGG)e (SEQ ID NO: 3) sequence, namely (CCG)e (SEQ ID NO: 4), and transfected, besides a control inhibitor, (by gymnosis, i.e. without transfection reagent) to four different ovarian cancer cell lines, ES2, TykNU, CAOV3, and OVCAR3, each in triplicates. Using the LNA based RT-qPCR systems, the expression of the GGA and the GGGA mirxons were examined and for both systems a downregulation was seen upon silencing of the (CGG)e (SEQ ID NO: 3) sequence, more pronounced for the GGGA mirxon (Figure 8). Figure 8 shows the RT-qPCR results for the two mirxon PCR systems (Figure 1) in four cancer cell lines upon silencing with QIAGEN miRCURY LNA miRNA Power Inhibitor for the (CGG)e (SEQ ID NO: 3) sequence, i.e. (CCG)e (SEQ ID NO: 4). Samples with “CGG” in the x- axis labels show the expression after silencing compared to controls with “CTRL” in the labels. The Y-axes in Figure 8 show Log2 expression (arbitrary unit). RNA-sequencing was performed and genes were identified which were associated with the inhibitor. 39 genes were significantly correlated with the inhibition over all four cell lines (FDR < 10 %), 36 (92.3 %) of them negatively (Figure 9A). Figure 9A shows a volcano plot of deregulated genes upon silencing of the (CGG)e (SEQ ID NO: 3) mirxons in four cancer cell lines. X-axis shows the Iog2 fold-changes between silenced cells and controls and the Y-axis the -logw p-value. The five most significant genes are labelled. Even more, looking on the Iog2 fold-changes a clear enrichment of negative Iog2 fold changes was revealed, involving hundreds of genes. Figure 10 shows the expression of the five most significant genes upon silencing with QIAGEN miRCURY LNA miRNA Power Inhibitor for the (CGG)e (SEQ ID NO: 3) sequence as determined by RNA-sequencing. The expression after silencing is shown in boxplots with “CGG” at the x-axis labels compared to controls with “CTRL” at the x-axis labels in Figure 10. The Y-axes in Figure 10 show the Log2 expression (arbitrary unit). For the first five highest significant genes (PPP3CA, ARID1 B, ERC1 , EIF4G3, and ANKS1 A, cf. Figure 10), all downregulated upon inhibition, the neighborhood of all first exons of all transcript variants were examined for (CGG)e (SEQ ID NO: 3) (CCG)e (SEQ ID NO: 4) repeats (Figure 11) and all showed at least for one first exon such repeats. Three of them had also one predicted mirxons and one even two in their first exons. Figure 11 shows the genomic loci of the five most significant deregulated genes upon silencing of the (CGG)e (SEQ ID NO: 3) mirxon. All of the genes in Figure 11 show (CGG)e (SEQ ID NO: 3) repeats in their first exons, three of them a predicted mirxon. DICER dependency of mirxon maturation and expression

[0282] Dicer, being part of the double-stranded RNA cleaving RNase III family, is involved in the maturation of miRNAs, i.e. it cleaves the pre-microRNA stem-loops into mature miRNAs. A study using the Dicer-negative fetal human kidney cell line 293T and small RNA sequencing (GSE146952) shows clearly that mirxon abundance is not affected upon removal of the DICER gene compared to the wild-type mother cell line, in contrast to the miRBase miRNA abundance reduction, indicating Dicer as not necessary for maturation of mirxons (Figure 12)

[0010] . In Figure 12 the Iog2-ratios of all mirxons up to five mismatches (“Mirxons”) and all miRBase miRNAs (“MiRBase”) to all miRNA reads (i.e. all species of RNAs implemented in the unitas v1.8.0 pipeline reported in the unitas.annotation_summary.txt-fi\e) are shown, split into samples derived from Dicerpositive (293T, left) and Dicer-negative (Dicer KO, right) cells.

[0283] Impact of mirxon expression on outcomes in cancer patients (clinical relevance)

[0284] Colorectal cancer: From 52 macro-dissected liver metastasis FFPE tissues of colorectal cancer patients, small RNA sequencing data were used to calculate the ratio of mirxon derived reads (CCGe up to five mismatches) related to all miRNA associated reads and used in a multiple Cox regression analysis for event free survival. All patients were operated for liver metastases as only recurrence or remaining tumor mass and event free survival was defined as time to next recurrence or progression. Splines were used to determine the optimal cutoff and the ratio was dichotomized according this result (Figure 13A). Figures 13B-C show the event free survival — colorectal cancer patients with liver metastasis. Figure 13A shows the splines of the hazard ratio of a Cox regression model correlated to the ratio of mirxon derived reads ((CGG)e (SEQ ID NO: 3) up to five mismatches) related to all miRNA associated reads (X-axis). The blue line in Figure 13A shows the optimal dichotomization value, 24 samples with low ratio versus 28 samples with a high ratio. A multiple Cox regression model was setup including relevant clinicopathologic parameters and the final model built using a step-wise downselection procedure, minimizing the Akaike Information Criterion (AIC). In Table 1 the relevant clinicopathologic parameters are shown. Figure 13B shows the Kaplan-Meier estimate for event-free survival of 52 colorectal cancer patients with liver metastases after resection of the liver metastases. Patients were divided into two groups according the ratio of mirxon associated reads compared to all miRNA associated reads from small RNA-sequencing data of macrodissected FFPE tumor tissues of the liver metastasis. The upper curve represents low mirxon expression and the lower curve high mirxon expression, respectively. Figure 13C shows the event-free survival curves of a Cox regression model including all relevant clinicopathologic parameters (Table 1 and Table 2). Again, upper curve, low expression, lower curve, high expression. As these survival curves are estimated from a multiple Cox regression model (HR 2.20 [CI95 1.12- 4.35]; p=0.023), no censored patients are seen. All significant confounding clinicopathologic parameters are shown in Table 2. Table 1 Clinicopathological parameters of patients of colorectal cancer with recurrences solely in the liver which were neoadjuvantly treated and operated.

[0285] Parameter Total Low High

[0286] N = 52 N = 24 N = 28 P-value1

[0287] Bilateral No 27 12 15 1.000

[0288] Yes 25 12 13

[0289] Tumor free Yes 10 4 6 0.736

[0290] No 42 20 22

[0291] No. of liver metastases High 15 6 9 0.760

[0292] Low 37 18 19

[0293] % Vital mets. tissue Low 13 7 6 0.541

[0294] (after neoadj. CHX)

[0295] High 39 17 22

[0296] BMI Low 16 6 10 0.549

[0297] High 32 18 18

[0298] 1Fischer's exact test

[0299] Table 2 Univariate and multiple Cox regression models for event free survival in colorectal cancer patients with liver metastases as solely recurrences.

[0300] N = 52 (42 events) Univariate Multiple2

[0301] Parameter HR1P-value HR P-value

[0302] Bilateral (yes vs. no) 1.05 [0.56-1.94] 0.883 0.40 0.038

[0303] [0.17-

[0304] 0.95]

[0305] Tumor free (yes vs. no) 0.19 [0.09-0.42] <0.001 0.08 <0.001

[0306] [0.03-

[0307] 0.21]

[0308] No. of liver mets. (high vs. low) 1.53 [0.79-2.95] 0.209 4.36 0.002

[0309] [1.73-

[0310] 11.03]

[0311] % Vital mets. tissue (high vs. low) 2.10 [1.00- 4.41] 0.051 4.29 0.001

[0312] [1.78-

[0313] 10.31]

[0314] BMI (high vs. low) 0.67 [0.35-1.28] 0.230 0.41 0.017

[0315] [0.20-

[0316] 0.85]

[0317] Mirxon (high vs. low) 1.75 ( 0.92- 3.32] 0.089 2.20 0.022

[0318] [1.12-

[0319] 4.35]

[0320] 1Hazard Ratio [95% Confidence interval],2Removed from the model by step-wise downselection, minimizing the Akaike Information Criterion (AIC) were the following additional clinicopathologic parameters: Age, Sex, Grading, No. of cycles of neoadjuvant chemotherapy, Diagnosis type (synchronous or metachronous), Type of intervention (primary versus recurrence), Diameter of liver metastasis.

[0321] Ovarian cancer I: The two OVARIAN mirxon RT-qPCR systems (GGA / GGGA, cf. Figure 1) targeting the 3’-ends of two mirxon-families defined by the 3’-end sequences given in Figure 1 (due to experimental constraints for the RT-qPCR the 3’- region has to be completely defined but the 5’-ends can be flexible and is therefore not coded into the oligos used for RT-qPCR) were used to examine the expression of these mirxon families in 30 fresh frozen ovarian cancer tissues (normalized to the geometric mean of three normalizer small RNAs, i.e. hsa-miR-191-5p, 5S rRNA, and a tRNA, Figure 14A) and the expression values were used to estimate the impact on overall survival. For both mirxon-families the optimal cutoffs for dichotomization in Cox regression analyses were estimated, i.e. 19 low versus 11 high for GGA-mirxons and 24 low versus 6 high for GGGA-mirxons. Finally, the dichotomized GGA and GGGA mirxon values were combined to three categories, both low (N=17), one high (N=9), and both high (N=4), respectively. A multiple Cox regression estimate including initially the combined GGA / GGGA expression value, age, FIGO-stage, and residual tumor after debulking surgery (R) revealed after feature selection an independent impact of the GGA / GGGA expression (both high, lower curve vs. one high, middle curve vs. both low, upper curve) on overall survival (HR 3.38 [CI95 1.65-6.91]; p<0.001 ; Table 3 and Figure 14C) in late stage high grade serous ovarian cancer (HGSOC) patients. The Kaplan- Meier estimates for the GGA / GGGA-mirxon expression are shown in Figure B, with the upper curve (red) for both low, the middle curve (green) one high, and the lower curve (blue) for both high. Mirxons GGA and GGGA (cf. Figure 1 and Figure 13A) were measured from isolated small RNA of fresh frozen tumor tissues by RT-qPCR and normalized to three small RNA normalizers (i.e. hsa-miR-191-5p, 5S rRNA, and a tRNA). Both mirxon expression values were optimally dichotomized for survival predictions (splines of a Cox regression model) and combined to a tripartite predictor (i.e. both high versus one high versus both low). Figures 14D-E show the overall survival — late stage high grade serous ovarian cancer via gene signature derived from silencing experiment (Upper curve (red), Signature high and lower curve (blue), Signature low).

[0322] Table 3 Clinicopathological parameters of late stage (FIGO lll / IV) high grade serous ovarian cancer patients.

[0323] Mirxons Total Both low One high Both high

[0324] Age 57.6 [49.7-67.3]157.1 [49.6-63.3] 66.6 [50.4-72.2] 57.6 [43.9-68.9] 0.857

[0325] FIGO III 22 13 6 3 0.848

[0326] IV 8 4 3 1

[0327] R 0 13 9 2 2 0.338

[0328] 1 17 8 7 2

[0329] 1Median [Interquartile range],2Linear regression or Fisher's exact test.

[0330] Table 4 Univariate and multiple Cox regression models for overall survival in late stage high grade serous ovarian cancer patients.

[0331] N = 30 (18 events) Univariate Multiple

[0332] Parameter HR1P-value HR P-value

[0333] Age (per decade) 1.96 [1.21-3.17] 0.006 2.21 [1.38-3.52] <0.001

[0334] FIGO (IV vs. Ill) 2.75 [1.05-7.18] 0.039 -2

[0335] R (Rl vs. R0) 1.46 [0.56-3.79] 0.438 2.85 [0.88-9.29] 0.082

[0336] Mirxons (GGA / GGGA) 2.40 [1.37-4.21] 0.002 3.38 [1.65-6.91] <0.001

[0337] 1Hazard Ratio [95% Confidence interval],2Removed from the model by step-wise downselection, minimizing the Akaike Information Criterion (AIC).

[0338] Ovarian cancer II: Using the (CGG)e (SEQ ID NO: 3) silencing results obtained from the cell culture experiment, a gene signature comprised of 297 higher-expressed genes in the (CGG)e (SEQ ID NO: 3) silenced samples and 268 higher-expressed genes in the control samples (CellMix, using the Abbas-method) was developed. In Figure 15 boxplots are shown using this 565 gene-signature applied to the cell culture samples, i.e. median expression of the 297 up-regulated genes minus median expression of the 268 down-regulated genes (gene-signature predictor) with silenced samples labeled with “CGG” and control samples with “CTRL” in the x-axis. In a next step, this genesignature predictor was applied to gene expression data of a comprehensive annotated cohort of 163 late-stage high grade serous ovarian cancer patients [3],

[0011] . Figure 14E shows the overall survival curves from a multiple Cox regression model of 163 high grade serous ovarian cancer patients divided into two groups according a predictive 565 gene signature derived from the (CGG)e (SEQ ID NO: 3) mirxon silencing experiment (for details see results and Figure 15). In Figure 14E survival curves from a multiple Cox regression model are shown, using the optimally dichotomized gene-signature predictor, corrected for age, FIGO stage, and peritoneal carcinosis (residual tumor mass was excluded from the model using a step-down selection procedure) - upper curve (red) for Signature High and lower curve (blue) for Signature Low. Optimal dichotomization was at 65 %, yielding 57 samples indicating the (CGG)e (SEQ ID NO: 3) silencing status, i.e. the 297 up-regulated genes are high expressed and the 268 down-regulated genes low, and 106 samples indicating high (CGG)e (SEQ ID NO: 3) mirxon expression. The 57 samples showed a favorable survival with a hazard ratio of 0.51 [CI95 0.31-0.86, p=0.011 ]. Figure 15 shows boxplots that show the difference of the 565 gene-signatures (comprised of 197 higher and 268 lower expressed genes in the (CGG)e (SEQ ID NO: 3) mirxon silenced cells), calculated as follows, median [297 higher] - median [268 lower] expressed genes. The Y-axis of Figure 15 shows the Log2 expression (arbitrary unit).

[0339] Ovarian Cancer III: To further validate the impact on overall survival in ovarian cancer, 16 publicly available gene expression studies were used as above, i.e. gene signature algorithm applied, optimally dichotomized, Cox regression estimated, and finally sample number-weighted summarized. In Figure 14D the hazard ratios for all of these 16 studies and the weighted summary are shown (sumHR 0.77 [CI95 0.59-1.00, p=0.05]). Figure 14D shows the validation of the 565 gene signatures in 16 publicly available gene expression data-sets (see above). As these results are correlated directly with the (CCG)e (SEQ ID NO: 4) silenced genes, the HR below one fits to-and therefore validates-the HR above one of the direct effects of the mirxon expressions on overall survival in HGSOC above.

[0340] Correlation of mirxons expression with a clinical and molecular subtype in HGSOC (clinical relevance)

[0341] A new molecular and clinical subtype was described, based on the peritoneal metastasizing pattern in HGSOC patients, i.e. many millet sized tumor nodules on the peritoneal wall, called miliary (“Mi”), versus only few large or no peritoneal metastases, called non-miliary (“nMi”) [4],

[0011] —

[0016] and correlated the ratio of the expression of mirxons (up to five mismatches) to all miRNAs from isolated and enriched tumor cells derived from tumor cell aggregates or single ascites tumor cells (“AS”) and primary or metastatic tumors (“PM”) [4], In Figure 16 the differences of the mirxon-to-all miRNA ratios between nMi and Mi tissues are shown, highly significant (p=0.004) and corrected for tumor tissue type. Figure 16 shows the Iog2 ratios of mirxon associated reads (mismatch up to five) to all miRNA associated reads in pure tumor cells (isolated by means of EpCAM-affinity beads after enzymatic dissociation of tumor tissues) of high grade serous ovarian cancer, subdivided into samples from single ascites tumor cells (A) or tumor spheroids (S) from ascites and primary (P) or metastatic tissue (M) and according peritoneal tumor spread types, miliary (Mi) and non-miliary (nMi). The difference between spread types was significant, corrected for tumor tissue type (p=0.004). The Y-axis of Figure 16 shows the Log2 expression (arbitrary unit).

[0342] Mirxons in extracellular vesicles (EVs) of cell free plasma of ovarian cancer patients and cell free semen (clinical relevance)

[0343] To assess the presence of mirxons in extracellular vesicles (EVs) like exosomes, which are known to contain microRNAs, raw small RNA sequencing data sets of EVs of cell-free semen (GSE203494

[0017] ) and cell-free plasma of ovarian cancer patients (GSE166310, unpublished) together with clinicopathologic parameters were downloaded from Gene Expression Omnibus GEO (https: / / www.ncbi.nlm.nih.gov / geo / query / acc.cgi?acc=GSE166310) and analyzed as above. The mirxon-ratio (reads associated with mirxons divided by reads associated to all microRNAs) was determined and correlated to the clinical data of the respective GEO data set. Ratios compared to those in Figure 6B and 6D from human tissues (up to six mismatches), shown in Figure 17A and 17B, were in the upper range. For both, the mirxon-ratio MM0_4 (up to four mismatches) was significantly associated with the clinical characteristics, i.e. statistical significant lower in infertile semen (p=0.031 , corrected for sperm concentration, motility, and morphology; 17C) and significant higher in EVs of plasma of none-RO respective ovarian cancer patients (p=0.021 ; 17D). Figure 17 shows the Iog2 ratios of mirxon associated reads in extracellular vesicles of (Figure 17A) cell-free semen of fertile and infertile males and (Figure 17B) cell-free plasma of ovarian cancer patients. Y-axes of Figure 17A and B show the Log2 ratio mirxons (with mismatches indicated at the x-axis) to all miRNAs. Figure 17C shows the correlation of the Iog2 mirxon expression ratio (up to four mismatches) with infertility status in 96 cell-free semen of males (corrected for concentration, motility, and morphology) and Figure 17D shows the difference of the Iog2 mirxon expression ratio (up to four mismatches) in cell- free plasma of ovarian patients, either resected RO or not. The Y-axes of the plots in Figure 17C are from left to right: Ratio, Concentration, Motility, Morphology.

[0344] Correlation of mirxons in extracellular vesicles of benign versus malignant tumor cells

[0345] Tumor cells from 22 early-stage head and neck squamous cell carcinoma (HNSCC) tissues and ten benign tumors as controls were cultivated for 48 hours in medium with exosome depleted FCS (GSE200489)

[0018] , RNA from extracellular vesicles was isolated and small RNAs sequenced. Reads were analyzed as described above and the ratios of reads associated with mirxons divided by all reads associated with microRNAs.

[0346] In Figure 18 (left) the distributions of these ratios (mirxons with zero (MMO) to six mismatches (MM6) and sums of mirxons with zero and one mismatch (MM0_1) to mirxons with zero to six mismatches (ALL6)) over all samples are shown. The Y-axis of Figure 18 (left) shows the Log2 ratio mirxons (with mismatches indicated at the x-axis) to all miRNAs. In Figure 18 (right) the significant difference of the ratios of MM0_4 (mirxons with zero to four mismatches) between benign and malignant (i.e. HNSCC) is shown. Figure 18 right panel: 0 refers to benign, 1 refers to malign. The two high values in the benign group are from tumor cells of a schwannoma and a pituitary adenoma.

[0347] Correlation of mirxons from extracellular vesicles of endometrial fluids with the implantation of a fertilized ovum in the endometrium during in vitro fertilization (clinical relevance)

[0348] RNA from extracellular vesicles (EVs) of endometrial fluids (EF) from 30 patients with implantative endometrium and from 30 patients with non-implantative endometrium (no implantation of the fertilized ovum happened after in vitro fertilization (IVF)) were isolated and small RNAs sequenced (GSE178917)

[0019] . Mirxon associated reads were analyzed as above and in Figure 19 the results are shown. On the left side the ratios of mirxon expressions over all samples and on the right side the significant differences between implantative and non-implantative endometrium of the ratios MM0_6 are shown. The Y axis of Figure 19 (left) shows the Log2 ratio mirxons (with mismatches indicated at the x-axis) to all miRNAs.

[0349] Correlation of mirxons from plasma of colorectal cancer patients and healthy controls (clinical relevance)

[0350] Circulating small EVs from plasma of 90 healthy controls and 120 colorectal cancer (CRC) patients (and further 80 samples from longitudinal follow-up time points) were analyzed as above (GSE188627)

[0020] . Figure 20 (left) shows the ratios of mirxon expressions over all samples and (right) the significant differences between healthy and colorectal cancer patients of the MM0_4 ratios. The Y-axis of Figure 20 (left) shows the Log2 ratio mirxons (with mismatches indicated at the x-axis) to all miRNAs.

[0351] Analyzing the miRNA read counts from this dataset with R-package limma, using function voomWithQualityWeights and cyclic-loess normalization, and building a discriminative logistic regression model (with elastic-net regularization, R-package glmnet) from the 30 most different expressed mirxons between healthy controls and CRC-patients revealed a predictor comprised of 11 mirxons with excellent discriminative power between healthy controls and CRC patients (model referred to as 11 -mirxon model). In Figure 21A the boxplots at time points 1 and 2 show the predictor values of the 11 -mirxon model of the 70 healthy controls and the 90 CRC patients, respectively, proving an excellent discriminative power. Furthermore, in Figure 21A the predictor values of the 11 -mirxon model are shown for one patient at 13 follow-up time points (time points when follow-up examinations are made with CRC patients, usually in 3-6 months intervals), line with open circles, and the tumor sizes, line with open triangles. In Figure 21 B the same is seen from the best discriminative miRBase miRNA hsa-miR- 320c

[0020] , line with closed circles, and tumor sizes, line with closed squares. The discriminative mirxon-model already increases two time points before the increase of the tumor size (blue arrows in Figure 21A at time points 8 and 9), whereas the best candidate miRBase miRNA (hsa-miR-320c) from Yang et al. 2022

[0020] publication did not indicate (arrows at time points eight and nine) the recurrence before measurable size increase at time point ten. This shows that the 11-mirxon model predictor indicates the recurrent disease already at least six months (probably 12 months, dependent on the follow-up interval) before the tumor size increase was recognized by clinical examinations. The best discriminative miRBase miRNA was not able to do this.

[0352] Differentiation of diffuse large B-cell lymphoma (DLBCL) from healthy controls Small RNAs were isolated from FFPE tissues of 17 healthy controls and 78 DLBCL tissues and sequenced (GSE185796)

[0021] , Raw sequencing reads were downloaded from the GEO Gene Expression Omnibus repository and were analyzed as above (i.e. Unitas-pipeline with mirxon annotations and R-package limma analysis with function voomWithQualityWeights and cyclic loess normalization). In Figure 22 the best discriminating miRBase miRNA hsa-miR-4733-3p and the best discriminating mirxon atlasMiR-106609-3p, a mirxon with three mismatches, between CTRL and DLBCL samples are shown. The Y-axes show the relative Iog2 expression (arbitrary unit). The differentiation between cases and controls is for the atlasMiR-106609-3p mirxon nearly perfect, with only one overlapping sample, compared to the miRBase miRNA with many overlapping samples between controls and DLBCL cases.

[0353] Tissue dependent correlation of miRNAs with age in mice

[0354] Small RNAs were isolated from mice tissues at different time points of the ontogenesis (GSE217458)

[0022] , Raw sequencing reads were analyzed as above (unitas pipeline with mouse-specific mirxons added and R-package limma). Shown in Figure 23 are three tissues with the best age-dependent miRBase miRNA mmu-miR-29c-3p (as described in the Wagner et al 2024

[0022] ) and the best age-correlated mirxon, mirxon- 1895-3p. The Y-axes of Figure 23 show the relative Iog2 expression (arbitrary unit). Clearly, the mirxon correlates much better with age than the miRBase miRNA. On the x-axis the age of the mice is given in weeks, from 1 to 27. The samples IY, HY, HA, and IA are unimportant here and can be neglected.

[0355] Correlation of mirxon expression to DNA-methylation of CpG islands and gene bodies

[0356] To correlate the relative mirxon expressions (i.e. the ratio mirxon derived reads per all miRNA-derived reads) to DNA methylation small RNA sequencing and reduced representation bisulfite sequencing (RRBS) data were used from FFPE tumor tissues of 36 liver metastases of colorectal cancer patients. Small RNA expressions were quantified using the unitas pipeline as above and DNA methylation of CpGs over CpG- islands (CGIs) and gene bodies (GBs) were calculated from RRBS data (weighted average p-values) as described in Pils et al.

[0023] These CGI and GB averaged p-values were correlated to the relative mirxon expression values (CCGe up to five mismatches) using linear regression models (R-package limma). Figure 24 and 22 show the distributions of the Iog2 fold-changes (FCs) of the CGIs and GBs revealing a clear shoulder at the right, i.e. positive Iog2 FCs, only for the CGIs (Figure 24) but not for GBs (Figure 25). In Figure 26 the Iog2 FCs are plotted against the median averaged p-values, showing the overrepresented Iog2 FCs in CGIs mainly for the low to medium averaged p-values. These results show a positive correlation of the mirxons expression with CGI methylation in a few hundred genes in cancer tumor tissues.

[0357] In Table 4, 32 analyzed datasets are summarized, indicating the clinical question (“Subject / Question”), the source of the small RNAs (“Tissue / Source”), the method (“Method”), the used measure for the evaluation of the mirxon expression (“Mirxon measure”), the source of the data (“Data set(s)” with reference), the number of samples used (samples), the question / outcome of the corresponding publication (“Question / Outcome”), and the (significant) results with the direction of the correlation (“Result” and “Direction”). The analysis procedure was always the same: downloading of data from the GEO, quality control, trimming and mapping of reads to all miRNAs (ENCODE) and other RNA species (unitas v1.8.0 pipeline) and calculation of the ratio mirxon associated reads to all miRNA associated reads. This ratio was correlated to the biological / clinical question raised in the paper associated to the data set or inferred from the GEO experiment description, either univariately (Kruskal-Wallis test) or using confounding parameters in a model (multiple logistic regression), if available.

[0358] GSE110381 , GSE188627, and GSE71008 (cf. Table 5) contain samples from colorectal cancer patients (malignant), colon adenoma patients (benign) and healthy controls. Using R-package glmnet (v4.1-8) and samples from GSE188627 a CRC specific model comprising 11 mirxons (CRC model) to discriminate malignant samples of healthy controls was built with an area under the Receiver Operating Characteristic (ROC) curve (AUC) of 0.916 and validated with corresponding samples from GSE71008 with an AUC of 0.698. The same model was not able to discriminate adenoma samples from healthy controls (AUC=0.514, underlined in Table 5). Using GSE110381 , a model comprised of 10 mirxons specific for the discrimination of adenoma samples from healthy controls was built with an AUC of 0.694. This model was not able to discriminate CRC samples from healthy controls, AUC for GSE188627 of 0.511 and for GSE71008 of 0.574 (both underlined in Table 5). These results shows that a selection of mirxons can be used to discriminate specific CRC patients from healthy controls and adenoma patients from healthy controls with nearly no cross-specificity.

[0359] For pig (sus scrofa, susScr3 genome) and mouse (mus musculus, mm10 genome) the same procedure with tissue atlas samples was performed, yielding 756 and 1 ,595 mirxons, respectively. Mirxons from mouse and pig tissues show a similar ratio (mirxon expression values to all miRNA expression values) distribution and tissue specificity as shown for human tissues (cf. Figures 6B and 6D).

[0360] Triplex helix potential of mirxons

[0361] To assess the potential of mirxons to build triplex helices RNA:dsDNA to its own place of origin in the genome, TriplexAligner

[0024] was used to simulate triplex helices with random sequences with itself, random sequences with other random sequences, all possible trinucleotide repeats, and all miRNAs with its own places of origin, all with length of 18 nucleotides. From the eight possible models trained for the TriplexAligner software, code 7 was used, as this code yielded the best biophysical validity upon comparison with RADICL-seq

[0025] pair alignments. Figure 27A shows the histograms of —logic E-values of the potential to build RNA:dsDNA triplexes from random RNA sequences to itself as dsDNA (green, right peak), or to another random sequence (cyan, left peak) and all trinucleotide repeats to itself (six violet bars in the right peak). All RNAs were of length 18. E-values were calculated with TriplexAligner using code 7 as model. The left (red) arrow below the x-axis indicates the E-value of a 19 nucleotides long RNA: dsDNA combination whose triplex helix was experimentally verified by the authors of TriplexAligner

[0024] , The right (violet) arrow below the x-axis indicates the E-value of the (CGG)e (SEQ ID NO: 3) RNA sequence building a triplex helix with its progenitor dsDNA- sequence. In Figure 27A the distributions of the —logic E-values for RNA:dsDNA helix building of 18 nucleotides long RNAs random to itself, random to random and all possible trinucleotide repeats are shown. E-values were calculated according Karlin-Altschul statistics and are length dependent. A red arrow below the X-axis shows the —logic E- value of the best candidate from the TriplexAligner paper which was experimentally verified

[0024] , The right (violet) arrow below the x-axis shows the - logic E-value of the (CGG) / GCG / GGC (and TGG / GTG / GGT) trinucleotide repeats. Figure 27B shows E- values of all possible RNAs with trinucleotide repeats (NNN)e (SEQ ID NO: 28) upon binding to their own dsDNA sequences. The (GGG)e (SEQ ID NO: 29) shows the highest E-value and the (CGG)6(SEQ ID NO: 3) and (TGG)e (SEQ ID NO: 30) the second highest one. (CGG)e (SEQ ID NO: 3) are the mirxons and TGG the most frequent mutations in mirxons. This value is indicated with the left (violet) arrow in A. In Figure 27B —logic E-values of all possible trinucleotide repeats are shown (combining all possible equivalent combinations, i.e. (CGG)=(CGG) / GCG / GGC, into one). Figure 27C shows E- values of all mirxons with mismatches up to five (“MM0_5”, brown (always the left boxplot of the boxplot triples for each length as indicated in the X-axis)), mirxons with six mismatches (“MM6”, red (always the middle boxplot)), and all other miRNAs (“No”, blue (always the right boxplot of the boxplot triples for each length)), grouping them into all possible length (E-values are dependent on length). As line only for length 18 the —logic E-value of the perfect (CGG)e (SEQ ID NO: 3) is shown (as given in Figure 27B). For all length the E-values are higher for mirxons (up to five mismatches and less pronounced for six mismatches) compared to other miRNAs.

[0362] able 5 Characteristics of 32 small RNA sequencing or gene expression datasets used for evaluation of mirxonifferences in clinical questions.

[0363] Subject / Question Tissue / Source Method Mirxon measure Source of Samples Question / Outcome Result Direction

[0364] (Human, if not Data set(s) other specified)

[0365] Colorectal cancer Macrodissected Small RNA Ratio own 52 Event free survival p=0.023 Higher worse

[0366] (CRC) FFPE sequencing

[0367] Ovarian cancer Fresh frozen RT-qPCR AACT (3 own 30 Overall survival p<0.001 Higher worse

[0368] (OvCa) tissues 2 Mirxons Normalizer)

[0369] OvCa Fresh frozen Microarray Via 565 gene- own 163 Overall survival p=0.011 Higher worse tissues signature GSE49997 [3]

[0370] OvCa Fresh frozen Various Via 565 gene- 16 Various Overall survival p=0.050 Higher worse tissues (RNA-seq, signature

[0371] Microarray) OvCa EpCAM Small RNA Ratio own 26; 21 Subclass (miliary vs. non- p=0.004 Miliary < enriched tumor sequencing miliary) - 574 mirxons non-miliary cells / fresh

[0372] OvCa Extracellular Small RNA Ratio GSE166310 14; 11 RO-resectability p=0.021 Higher nonvesicles (EVs) sequencing (unpublished) RO plasma: R1 vs. RO Semen: Infertiliy EVs of cell free Small RNA Ratio GSE203494 59; 32 Fertility versus infertile p=0.031 Higher semen sequencing

[0017] infertile

[0373] Implantative EVs of Small RNA Ratio GSE178917 30; 30 Non-implantative or p=0.048 Higher

[0374] Endometrium endometrial sequencing

[0019] implantative endometrium implantative fluid Head and neck EVs of cell Small RNA Ratio GSE200489 22; 10 Malignant versus benign p=0.006 HNSCC > squamous cell culture sequencing

[0018] benign carcinoma medium

[0375] Colon adenoma EVs from Small RNA Ratio GSE110381 383; 439 Adenoma versus healthy p=0.020 CTRL > plasma sequencing

[0026] control Adenoma

[0376] Adenoma model 10 383; 439 Adenoma versus healthy AUC=0.694 Adenoma mirxons control model 10 mirxons Validation: CRC 383; 439 Validation CRC model AUC=0.514 CRC model 11 model 11 mirxons mirxons

[0377] CRC EVs from Small RNA Ratio GSE188627 120; 90 CRC versus healthy p=0.005 CTRL > plasma sequencing

[0020] Cancer

[0378] CRC model 11 120; 90 CRC versus healthy AUC=0.916 CRC model 11 mirxons mirxons

[0379] Validation: 120; 90 Validation adenoma model AUC=0.511 Adenoma

[0380] Adenoma model 10 model 10 mirxons mirxons

[0381] CRC; Prostate EVs from Small RNA Ratio GSE71008 50; 100; Cancer versus healthy p=0.003 CTRL > PrCa; cancer (PrCa); plasma sequencing

[0027] 36; 6 CTRL > PDAC

[0382] Pancreatic cancer Ratio 192 Age p<0.001 Higher older

[0383] (PDAC) (>55 yrs)

[0384] Validation: CRC 50; 100 Validation model CRC AUC=0.698 CRC model 11 model 11 mirxons mirxons

[0385] Validation: 50; 100 Validation model Adenoma AUC=0.574 Adenoma

[0386] Adenoma model 10 model 10 mirxons mirxons

[0387] Ratio 274 Pre versus post training p<0.001 Higher post

[0388] (both conditions) training

[0389] Cardiomyopathy Tissue Small RNA Ratio GSE53080 17; 21; 30 Healthy vs. 1 CM vs. DCM p<0.014 DCM > ICM > sequencing

[0028] Healthy

[0390] Plasma Ratio 17; 46; 22 Healthy vs. ICM vs. DCM p<0.043 DCM > ICM;

[0391] DCM > Healthy

[0392] Lung Fetal and adult Small RNA Ratio GSE175462 25; 77; 63 Fetal vs. adult liver vs. p<0.001 Adult > fetal; adenocarcinoma lung and sequencing

[0029] carcinoma ca > fetal carcinoma Ratio 40; 100 Male (M) vs. female (F) p=0.033 Male >

[0393] Female Ratio 53; 41; 46 Smoker vs. ex-smoker vs. p<0.002 Non > ex > non-smoker smoker

[0394] Ratio 45; 113 Third vs. first trimester p=0.049 First > third trimester

[0395] T cell Cells Small RNA Ratio GSE181216 25; 24 CD4 vs. CD8 T cells p<0.001 CD8 > CD4

[0396] EVs within the sequencingRatio

[0030] 34; 27CD4 vs. CD8 T cells p<0.001 CD8 > CD4 immunological synapse Alcohol use Post-mortem Small RNA Ratio GSE181982 442; 369 Alcohol use disorder vs. p=0.017 AU D > CTRL disorder; brain brain biopsies sequencing

[0031] control

[0397] (8 different Ratio 554; 69; Smoker vs. ex-smoker vs. p<0.001 Non > regions) 188 non-smoker smoker

[0398] Martial arts: mild Plasma and Small RNA Ratio GSE123336 89; 129 Pre; post (up to 65 hits p=0.002 Higher more traumatic brain saliva sequencing

[0032] against the head) hits injury Ratio 87; 131 Saliva; serum p<0.001 Saliva > serum

[0399] Stool: celiac Cell free stool Small RNA Ratio GSE217915 66; 3; 50; Healthy; CD; treatedCD-TG'; p<0.025 F: CD < CTRL disease sequencing

[0033] 13 tCD-TG+M: tCD-TG+< CTRL

[0400] Nociceptor Three cell lines; Small RNA Ratio GSE161530 54 Three cell lines (840, AD2, p<0.001 Lower ratio maturation Six sequencing

[0034] AD3); Six developmental with higher developmental states (iPSC -> nociceptor) maturation states state

[0401] Lung disease: ILD, Tissue biopsies: Small RNA Ratio GSE201121 66; 152; Healthy vs. ILD vs. COPD p<0.001 ILD > COPD > COPD Healthy; ILD; sequencing

[0035] 153 Healthy

[0402] COPD

[0403] Age Cell free Small RNA Ratio GSE182598 64 <65 <65 yrs vs. >65 yrs p<0.001 Higher ratio plasma; 91 sequencing

[0036] yrs, in <65 yrs male and 12 39 >65 female; age 20- yrs; 83 yrs, IQR 91 male, [45.5 68.5] 12 female

[0404] Primary cell line 38 cell types Small RNA Ratio ENCODE 38 Expression in cell types - Log2 ratios: - sequencing 1.113 mirxons 11 to -4

[0405] Primary tissue 36 cell types Small RNA Ratio ENCODE 82 Expression in 36 different Log2 ratios: - sequencing human tissues 11 to -4

[0406] RNA Atlas Primary tissues, Small RNA Ratio GSE138734 180; 636; Expression in primary Log2 ratios: - (Tissue, Cell type, cell types, cell sequencing

[0037] 20; 350 tissues, cell types, 14 to -4 Cell line, Cancer lines, cancer (cancer) cell lines cell line) cell lines

[0407] DICER knockout HEK293T Small RNA Ratios GSE146952 8 vs. 12 DICER necessary for p<0.001 DICER sequencing

[0010] maturation of mirxons dependent

[0408] AG02 pull-down HeLa: AGO2 IP- Ratios GSE43666 I vs. I vs. mirxons are depleted in n.d. AGO2 asynchronous seq

[0038] CTRL AGO2 immunoprecipitates independent? and mitotic cells

[0409] AG02 and HuR HEK293 AGO2 CLIP Ratios GSE28865 6 vs. CTRL mirxons are depleted in n.d. AG02

[0410] (HuA) (PAR)-CLIP and PAR-

[0039] AG02 CLIP independent?

[0411] CLIP Mouse: Tissue Mus musculus Small RNA Ratios GSE217458 947 tissue Tissues, EV-enriched, free- p<0.001 Logz ratios: - and age atlas; EV- Tissue biopsies sequencing

[0022] samples; circulating (different ages) - 12 to -5 enriched and 56 EV- 1.595 mirxons free-circulating enriched and free- circulating Mouse: Microglia Mus musculus Small RNA Ratios GSE229983 70 tissue Microglia (CD45+) vs. p=0.004 CD45+ >

[0412] (CD45+) and Tissue biopsies sequencing

[0040] samples; other cells (CD45-) p<0.001 CD45- other cells (CD45- 35 CD45+ P6 vs. P15 vs. week 8 p<0.001 CD45+: P6 >

[0413] ), different time and 35 P15 > w8 points CD45-; 22 CD45-: P6 <

[0414] P6, 24 P15 < w8

[0415] P15 and 24 week 8 Pig: Tissue atlas Sus scrofa Small RNA Ratios GSE162147 187 tissue 71 Tissues - 756 mirxons p<0.001 Logz ratios: -

[0416] Tissue biopsies sequencing

[0041] samples 18 to -6

[0417] Area under the Receiver Operating Characteristic (ROC) curve (AUC); not determined (n.d., highlights that mirxons are tissue specific); bold: significant results; italic: not significant; underlined: model validations.

[0418] Conclusion

[0419] As shown above, mirxons demonstrate better discriminative characteristics compared to other small RNAs in many biological and clinical questions.

[0420] Example 2

[0421] The following example highlights that mirxons are missing from prominent miRNA databases.

[0422] In the human genome HG38 that is described herein in example 1 , 5,622 nonoverlapping mirxons with their own positions on the genome, mostly in first exons or promoter regions of other genes. The mismatch distribution in the core-region (CGG)e (SEQ ID NO: 3) or (CCG)6(SEQ ID NO: 4) is shown in Table 6.

[0423] Table 6 Mismatch distribution of mirxons in the human genome HG38.

[0424] Mismatch(es) 0 1 2 3 4 5 6

[0425] Numbers 235 258 259 394 416 950 3110

[0426] Available databases do not contain information on mirxons:

[0427] MiRBase v22.1 (https: / / www.mirbase.org / , accessed 08 / 05 / 2024)

[0042] has 0 mirxons (CGG)e (SEQ ID NO: 3) with 0 or 1 mismatches out of 2,656 human miRNAs, whereas there are 235 and 258 mirxons (sum: 493) present in the human genome with 0 and 1 mismatches, respectively.

[0428] Moreover, there are 0 mirxons with 0 or 1 mismatches out of all 48,885 miRNAs in the miRBase v22.1 , covering all 271 species from miRBase v22.1.

[0429] MirCarta v1.1 (https: / / mircarta.cs.uni-saarland.de / , accessed 08 / 05 / 2024)

[0043] , covering much more mature human miRNAs, namely 24,796, has two mirxons with 0 mismatches, out of the 235 known ones (corresponding to 0.9 %). Table 7 shows these two mirxons

[0430] Table 7 Mirxons with 0 or 1 mismatch included in the MirCarta v1.1 database. Both positions do not overlap with the 5622 non-overlapping mirxons from human genome HG38.

[0431] Acc.-No Species Type Position Sequence m-2669 Homo miRNA chr!9:36, 141, 037-36, 141, 060 AGGGCGGCGGCGGCGGCGGCGGGG sapiens (+) (SEQ ID NO: 21) m-2685 Homo miRNA chr2:26, 692, 715-26,692,737 CGGCGGCGGCGGCGGCGGCCCCG sapiens (+) (SEQ ID NO: 22) The present example thus highlights that mirxons have so far not been regarded as biologically relevant and have not been included in or have even been removed fromprominent databases.

[0432] Example 3

[0433] Binding of mirxons to genomic DNA.

[0434] Pull-down experiment proving in vivo the binding of the mirxon-oligo (SEQ ID NO: 23) to the genomic DNA in the nucleus. Briefly, a 20 nt long RNA-oligo with a (CGG)e (SEQ ID NO: 3) sequence or a random sequence (NEG) and psoralen (Pso) and biotin (Bio) on the 5’ end 3’ end, was transfected into cells followed by the isolation of cell nuclei. The nuclei were exposed to UV irradiation to crosslink the RNA-oligo to the genomic DNA via the psoralen. The sheared DNA fragments were purified using 100 pl pMACS streptavidin MicroBeads (Miltenyi Biotec) and pMacs columns, which capture the biotin-labeled RNA-DNA complexes through streptavidin (Strep) binding. After DNA isolation, NGS libraries were prepared and sequenced. The sequenced reads were mapped to the reference genome, identifying the locations of the mirxon binding-sites to the genomic DNA.

[0435] The mirxon-oligo is 5’-GGC GGC GGC GGC GGC GGC GG- 3’ (SEQ ID NO: 23), wherein the 5’ end is linked to psoralen and the 3’ end is linked to biotin.

[0436] Control oligo (NEG): 5’-TGT AAC ACG TCT ATA CGC CC- 3’ (SEQ ID NO: 24), wherein the 5’ end is linked to psoralen and the 3’ end is linked to biotin.

[0437] The ovarian cancer cell line ES-2 (RRID:CVCL_3509) was used

[0044] , Transfection of oligos with Pepmute and cultivation was performed for 6 hours.

[0438] If the oligo is bound to the DNA (e.g. as RNA-dsDNA triple helix or as R-loop) UV- light can crosslink both molecules.

[0439] After the six hours incubation, cells were trypsinized and harvested and the nuclei were isolated with a nuclei extraction buffer. The washed nuclei were cross-linked in ice- cold cross-linking buffer with UV light (365 nm) and lysed with Proteinase K digestion (0.1 mg / ml; 30 minutes) and sonicated with the UP200St ultrasonic processor at maximum Watt and the 100 % cycle for five minutes. DNA-shearing was performed in microtube AFA Fiber Snap-Cap (PN 520045) with the S220 Focused-ultrasonicator from Covaris. After removal of the debris (centrifugation at full speed, 4 °C for five minutes) the biotin-labeled DNA-fragments were isolated by pMACS streptavidin MicroBeads and the pMACsTM separator in combination with p columns. After washing the columns, the beads with the bound DNA-fragments were eluted from the column and the DNA precipitated with 10 % 2 M NaCI and 2.5* ice-cold 100 % EtOH. The DNA yield after pulldown experiment with the mirxon-oligo GGC GGC GGC GGC GGC GGC GG (SEQ ID NO: 23) and the control NEG oligo (SEQ ID NO: 24), specifically the DNA yield in ng of the CGG were ~60 ng and the NEG oligo <10 ng, respectively.

[0440] Using the CGG and NEG DNA samples from above and a shared control DNA (input control), sequencing libraries were generated using the NEBNext UltraTM II DNA Library Prep Kit for Illumina (New England BioLabs, USA), following the manufacturer's instructions and the libraries sequenced (2 x 50 bp paired-end). Obtained reads were mapped to the genome and analyzed using the Chip-ap pipeline

[0045] .

[0441] It could be shown that only the mirxon-oligo (SEQ ID NO: 23) enriched sequences were bound at promoter regions of genes (at the transcription start site) but not the NEG control oligo (SEQ ID NO: 24) or the input DNA. The mismatch status of the (CGG)e (SEQ ID NO: 3) sequence at the promoter region correlated negatively with the enrichment of DNA-fragments and thus binding of the mirxon-oligo: Fewer mismatches more binding and no CGG-sequence at all, no binding.

[0442] Change of chromatin accessibility by silencing or overexpression of the (CGG)e (SEQ ID NO: 3) mirxon

[0443] Methods

[0444] To see the differences in chromatin accessibility and RNA expression upon usage of different concentrations of transfected mirxon RNA-oligos a dose-response experiment was performed. Two different RNAs (Table 8: CGG (SEQ ID NO: 25) and a- CGG (SEQ ID NO: 26)) and one random RNA (Table 8: NEG (SEQ ID NO: 27)) for control were ordered (Microsynth, Austria) and used for transfection. Ribonucleic acid oligos were ordered with three sequences listed in Table 8. One represented the mirxons with the canonical sequence (CGG)e (SEQ ID NO: 3) which is called mirxon-RNA-oligo (CGG), another one represented the antisense RNA (a-CGG) which is complementary to (CGG)e (SEQ ID NO: 3) like an anti-miR, and the third RNA-oligo had a random sequence not present in the human genome as a control (NEG). For higher stability against degradation, the first and last bases were DNA bases.

[0445] Table 8 Overview of the RNA used for transfection

[0446] Name SEQ ID NO Sequence Concentration [pM]

[0447] CGG 25 CGG-CGG-CGG-CGG-CGG-CGG-CG 100 a-CGG 26 CGC-CGC-CGC-CGC-CGC-CGC-CG 100

[0448] NEG 1 TGU-AAC-ACG-UCU-AUA-CGC-CC 100

[0449] For the experiments, ES-2 cells were transfected with these oligos as above and cultivated for three days, then trypsinized and counted with the LUNA-FL™ Dual Fluorescence Cell Counter (Logos Biosystems, South Korea). Eight T25 culture flasks were seeded with 3.3 x 105cells / ml (12 % confluence) for ES-2 in total. After 24 hours, the media were changed, and 3 ml of new growth medium was added. The cells were transfected with different concentrations of RNA-oligos and transfection agent (Table 9) in 300 pl transfection buffer. After 15 minutes, the transfection complex was added dropwise to the cells.

[0450] Table 9 Overview of the used concentrations for transfection.

[0451] RNA TYPE SEQ ID NO RNA pM pL RNA TOTAL NM pL PEPMUTE

[0452] RNA

[0453] CGG 25 1 3 1 6.7

[0454] CGG 25 10 3 10 7.7 a-CGG 26 1 3 1 6.7 a-CGG 26 10 3 10 7.7

[0455] NEG 1 10 3 10 7.7

[0456] Precipitation of DNA / RNA for the dose-response experiment

[0457] After 48 hours, ES-2 cells were harvested by trypsinizing. The cell pellet was washed with ice-cold 1 x PBS twice and finally resuspended in 1 ml ice-cold 1 x PBS. Cells were counted by the CASY cell counter and analyzer using the 150 pm capillary by mixing 100 pl of the cell suspension with 10 ml CASYton. After measurement, 50,000 cells were used for Assay for Transposase-Accessible Chromatin using sequencing (ATAC-seq) by following the guidelines for sample submission for ATAC-seq instructions. Results

[0458] Dose-dependent chromatin accessibility changes induced by RNA-oligos in ES- 2 revealed by ATAC-seq

[0459] In the following experiments, the potential association between the methylation state, chromatin openness dependent on the transfected RNA-oligos (CGG (SEQ ID NO: 25), a-CGG (SEQ ID NO: 26), and NEG (SEQ ID NO: 27), cf. Table 9) and the ability of the mirxons to bind DNA was investigated. To study chromatin accessibility patterns, ATAC-seq was performed, a method that uses a hyperactive mutant Tn5 Transposase, to insert sequencing adapters into the open chromatin regions. It was hypothesized that the anti-mirxon RNA-oligo (a-CGG, SEQ ID NO: 26) competes with the mirxon RNA- oligo (CGG, SEQ ID NO: 25), inhibiting its binding and reducing chromatin openness at specific loci and that the mirxon RNA-oligo (SEQ ID NO: 25) could open some other (initially more closed) chromatin loci, whereas the NEG-oligo (SEQ ID NO: 27) show no effect on chromatin openess of any loci. To test this, three RNA-oligos (CGG (SEQ ID NO: 25), a-CGG (SEQ ID NO: 26), and NEG (SEQ ID NO: 27)) were transfected at different concentrations (1 nM and 10 nM) into cell lines ES-2 in a dose-response experiment. This is aimed at investigating a potential concentration-dependent correlation between these oligos and chromatin openness. For bioinformatical analysis, a ChlP-AP pipeline for unbiased ChiP-seq analysis was employed

[0045] . Sample replicates were merged for deepTools analyses and genome browser inspections for quality control. Results exhibited the typical “histone laddering” pattern, indicating successful tagmentation and library generation. Peaks in the obtained mapped sequencing reads for ATAC-seq indicated accessible chromatin regions.

[0460] Analysis of (CGG)e (SEQ ID NO: 3) sequence mismatches in CpG-islands with chromatin openness

[0461] To investigate the relationship between (CGG)e (SEQ ID NO: 3) sequence mismatches in CpG-islands and chromatin accessibility, a similar analysis was conducted as done with the pulldown experiment. In this study, the focus was on the relationship between sequence mismatches and chromatin openness, rather than analyzing the concentration effect. The CpG-islands were classified using the (CGG)e (SEQ ID NO: 3) sequence mismatches (MM) and were correlated to the chromatin openness as determined by ATAC-seq. A correlation of (CGG)e (SEQ ID NO: 3) sequence mismatches with chromatin openness was observed with any used RNA-oligo and concentration used for transfection, was analyzed. There was always a highly significant negative correlation between chromatin openness and (CGG)e (SEQ ID NO: 3) mismatch class (defined with the lowest mismatch (CGG)e (SEQ ID NO: 3) sequence in the CGI) in the CGI, with a complete lack of signal for CGIs without any (CGG)e (SEQ ID NO: 3) sequence. These findings suggest that CGIs containing a (CGG)e (SEQ ID NO: 3) sequence are more open and accessible for gene expression and even more, that the lowest mismatch (CGG)e-sequence in the CGI correlates with the chromatin openness, i.e. CGIs with (CGG)e-sequences (SEQ ID NO: 3) with 0 mismatches with the most open chromatin and CGIs with (CGG)e-sequences (SEQ ID NO: 3) with 6 mismatches with least open chromatin, and consecutive decreasing openness inbetween.

[0462] Relationship between the CpG-island characteristics, pulldown, and ATAC-seq

[0463] To challenge the argument that the strength of a CpG-island is primarily determined by its CpG content, the next analysis focused on determining whether mismatches have a stronger influence compared to other characteristics, such as CpG dinucleotide content. The relationship between CpG-island or promoter region characteristics and DNA methylation, ATAC-seq results, and pulldown assay results, were examined using multiple linear regression modeling using natural splines for the independent variables. This analysis aimed to identify parameters correlating with chromatin accessibility and pulldown assay efficiency. The predicting factors included CGI / promoter region length, CG content, CpG-dinucleotide frequency, TpG-dinucleotide frequency, and the number of mismatches in the (CGG)e (SEQ ID NO: 3) sequence with the fewest number of mismatches per region. Regions without a (CGG)e (SEQ ID NO: 3) sequence at all, were defined with a mismatch number of seven. A statistically significant strong correlation was observed between the minimal number of mismatches (“minMM”) in (CGG)e (SEQ ID NO: 3) sequences and ATAC-seq values. An analysis indicated that minimal mismatches explain 62 % of the variance in ATAC-seq values (R2= 0.25), with the complete model having an R2of about 0.4. R2, a measure of the goodness of fit of the correlation, indicates how well the independent variables predict the dependent variable, with a value of one representing a perfect fit. Similarly, pulldown assay results showed a strong negative correlation with minimal mismatches in (CGG)e (SEQ ID NO: 3) sequences, with an R2of 0.55 for the complete model and 0.45 (82 %) for minMM alone. These findings suggest that minimal mismatches in (CGG)e (SEQ ID NO: 3) sequences play a critical role in predicting chromatin accessibility and pulldown assay efficiency.

[0464] To further investigate the influence of minimal mismatches in (CGG)e (SEQ ID NO: 3) sequences on DNA methylation, the relationship between mismatches and the mean methylation value (beta-value) across all tissues from the human methylation atlas was analyzed. The impact of CpG-island / promoter region characteristics on the mean methylation value was analyzed. The univariate R2 values indicated that minimal mismatches and CpG frequency are equally important predictors of CGI methylation, each explaining about 69 % (R2= 0.38) of the variance of the complete model (R2of about 0.55). The analysis revealed a statistically significant strong positive correlation between the number of mismatches and mean methylation levels. The results suggest that CpG-islands with more mismatches in the (CGG)e (SEQ ID NO: 3) sequences are more methylated and less accessible.

[0465] To complete the analysis, the correlation between chromatin accessibility (ATAC- seq) and mirxon binding (pulldown experiment) was modeled. Log2 ATAC-seq values were used as the independent variables and natural splines of Iog2 pulldown values along with all CGI / promoter region characteristics as the dependent variables. A strong positive correlation was observed between pulldown and ATAC-seq values, indicating that higher pulldown values are associated with higher ATAC-seq values. All other factors were also significant (p-value < 10'16). The high R2value for pulldown suggests it is a strong predictor of ATAC-seq variability. The correlation between mismatches of (CGG)e (SEQ ID NO: 3) sequences and the ATAC-seq values reversed compared to the usual correlation, likely due to a stronger correlation between pulldown and mismatches. These findings highlight the relationships between chromatin accessibility, mirxon binding, and sequence characteristics.

[0466] Analysis of concentration effects on chromatin openness and mirxon binding

[0467] The general analysis of the relationship between mismatches in the (CGG)e (SEQ ID NO: 3) sequence of CpG-islands and chromatin openness being completed, next the effects of different RNA-oligo concentrations on chromatin openness were examined. It was analyzed whether chromatin openness correlated with mirxon RNA-oligo binding in a concentration-dependent manner. ATAC-seq were compared with pulldown results in the ES-2 cell line. A clear dose-dependent correlation in all experiments was observed.

[0468] The transfection of the CGG mirxon sequence (SEQ ID NO: 25) revealed 3,778 loci (most of them promoter regions) with dose-dependent increasing in chromatin openness compared to the NEG (SEQ ID NO: 27) control, whereas transfection of the a-CGG anti-mirxon sequence (SEQ ID NO 26) revealed 2,699 loci with dose-dependent decreasing in chromatin openness compared to the NEG (SEQ ID NO: 27) control. Interestingly, the CGG mirxon (SEQ ID NO: 25) opened loci were rather closed in the control sample and vice versa the a-CGG anti-mirxon (SEQ ID NO: 26) closed loci were rather open in the control sample. This shows clearly, that the mirxon expressions are positively correlated to more open chromatin at many genomic promoter regions initially closed and that after silencing of these mirxons by transfected anti-mirxons many genomic promoter regions get more closed which were initially more open.

[0469] Example 4

[0470] Biosynthesis pathway of mirxons

[0471] To examine the biogenesis of mirxons, the dependency on Drosha, Dicer, and AGOs was analyzed using miRNAs from miRBase as the positive control compared to mirxons. To assess Drosha dependency, the expression levels of miRNAs in a Drosha depleted cell line were compared to a wildtype cell line. These miRNAs were predominantly downregulated (negative Iog2 fold-change) when Drosha was depleted. The similar up-to-downregulation ratio between mirxons and miRBase miRNAs suggests that mirxons depend on Drosha in a similar way as miRBase miRNAs do.

[0472] The impact of sequential AGO enzyme elimination in different cell lines was assessed by knocking out AG01 and AG02 using CRISPR, and silencing AG03 with siRNA. If AGO genes were not involved, the down-to-upregulated gene ratio would be 0.5. The results indicated that miRNAs from miRBase depend on AGO, as the proportion of downregulated miRNAs increased with the removal of more AGO genes. In addition, the percentage of significantly deregulated genes increased from 3.5 % to 22.7 % with the successive AGO gene removal. In contrast, mirxons and other non-miRBase predicted miRNAs showed no such dependency or increase in deregulated genes. This suggests that miRBase miRNAs require AGO for stability and protection, whereas mirxons do not. This indicates that mirxons are not part of the RISC complex. The Iog2 ratio of miRNA reads to total reads in the human embryonic kidney (HEK) 293T cell line and Dicer knock-out cells (GSE146952) indicated that Dicer silencing leads to a marked reduction in miRNAs from miRBase, while mirxons were unaffected. These findings suggest that mirxons are independent of Dicer (Figure 12).

[0473] Taken together, these findings indicate that mirxons rely solely on Drosha for their biogenesis and do not require Dicer or AGO proteins, in contrast to the canonical miRNA pathway.

[0474] The above results show that mirxons are not in the RISC complex but are bound, as shown above, to the DNA directly. Therefore, some small RNA analysis methods would not register mirxons (e.g. the LEXOGEN® TraPR Small RNA Isolation Kit). Further, the function of mirxons differs from any other class of small RNAs:

[0475] • Localization in promoter regions or first exons of (protein-coding) genes and often in CpG-islands

[0476] • Direct binding of mature mirxons to the DNA at their place of origin

[0477] • Not present in the RISC complex

[0478] • DICER not involved during maturation

[0479] • Positive regulation of gene expression at the transcriptional level (therefore direct impact on gene expression if silenced or overexpressed)

[0480] Therefore, mirxons can be regarded as a new class of small RNAs with specific characteristics not so far described in the state of the art. Specifically, mirxons differ from the use as a single small RNA of known classes (MicroRNAs (miRNAs), Small Interfering RNAs (siRNAs), Piwi-lnteracting RNAs (piRNAs), Small Nucleolar RNAs (snoRNAs), or tRNA-Fragments) or with known functions.

[0481] REFERENCES

[0482] [1] C. Zhao etal., ExRNA, 2019, vol. 1 , no. 1 , p. 38

[0483] [2] M. Kuhn, J Stat Softw, 2008, vol. 28, no. 5, pp. 1-26

[0484] [3] D. Pils et al., Cancer Sci, 2012, vol. 103, no. 7, pp. 1334-1341

[0485] [4] A. Bachmayr-Heyda et al., Oncotarget, 2016, vol. 7, no. 26, pp. 39640-39653

[0486] [5] F. Abascal et al., Nature, 2020, vol. 583, no. 7818, pp. 699-710

[0487] [6] V. Ruhela et al., Frontiers in Bioinformatics, 2022, vol. 2

[0488] [7] Y. Luo et al., Nucleic Acids Res, 2020, vol. 48, no. D1 , pp. D882-D889

[0489] [8] D. Gebert et al., BMC Genomics, 2017, vol. 18, no. 1 , p. 644

[0490] [9] A. Kozomara et al., Nucleic Acids Res, 2019, vol. 47, no. D1 , pp. D155-D162

[0491]

[0010] Q.-J. Luo et al., Nat Commun, 2021 , vol. 12, no. 1 , p. 3397

[0492]

[0011] C. Bekos et al., Cancers (Basel), 2019, vol. 11 , no. 5, p. 698

[0493]

[0012] K. Auer et al., Cancer Lett, 2017, vol. 388, pp. 158-166

[0494]

[0013] K. Auer et al., Oncotarget, 2015, vol. 6, no. 19, pp. 17261-17275

[0495]

[0014] K. Auer et al., Oncotarget, 2016, vol. 7, no. 38, pp. 61336-61354

[0496]

[0015] S. Aust et al., Sci Rep, 2017, vol. 7, no. 1 , p. 42929

[0497]

[0016] A. Bachmayr-Heyda et al., Clinical Cancer Research, 2017, vol. 23, no. 8, pp. 2081-2092

[0498]

[0017] O. A. Oluwayiose et al., Environmental Pollution, 2023, vol. 329, p. 121529

[0499]

[0018] C. R. Galiveti et al., Head Neck, 2023, vol. 45, no. 1 , pp. 212-224

[0500]

[0019] J. Ibanez-Perez etal., Human Reproduction, 2022, vol. 37, no. 10, pp. 2375-2391

[0501]

[0020] C.-K. Yang et al., Cellular Oncology, 2022, vol. 45, no. 4, pp. 621-638

[0502]

[0021] A. Larrabeiti-Etxebarria et al., Sci Rep, 2023, vol. 13, no. 1 , p. 12159

[0503]

[0022] V. Wagner et al., Nat Biotechnol, 2024, vol. 42, no. 1, pp. 109-118

[0504]

[0023] D. Pils et al., Cancers (Basel), 2020, vol. 12, no. 8, p. 2257

[0505]

[0024] T. Warwick et al., Brief Bioinform, 2022, vol. 23, no. 6, p. bbac445

[0506]

[0025] A. Bonetti et al., Nat Commun, 2020, vol. 11 , no. 1 , p. 1018

[0507]

[0026] B. S. Roberts etal., Clinical Cancer Research, 2018, vol. 24, no. 9, pp. 2092-2099

[0508]

[0027] T. Yuan et al., Sci Rep, 2016, vol. 6, no. 1, p. 19413

[0509]

[0028] K. M. Akat et al., Proceedings of the National Academy of Sciences, 2014, vol. 111 , no. 30, pp. 11151-11156

[0510]

[0029] D. E. Cohn et al., Cancers (Basel), 2021 , vol. 13, no. 11 , p. 2686

[0030] P. F. Cespedes et al., Nat Commun, 2022, vol.13, no.1, p.3460

[0511]

[0031] Y. Lim etal., Transl Psychiatry, 2021, vol.11, no.1, p.504

[0512]

[0032] D. LaRocca etal., PLoS One, 2019, vol.14, no.1, p. e0207785

[0513]

[0033] A. Francavilla et al., Gut Microbes, 2023, vol.15, no.1

[0514]

[0034] M. Zeidler etal., Advanced Science, 2021, vol.8, no.21, p.2102354

[0515]

[0035] A. B. Pavel etal., Sci Rep, 2023, vol.13, no.1, p.13076

[0516]

[0036] J. Salignon etal., Aging, 2023, vol.15, no.12, pp.5240-5265

[0517]

[0037] L. Lorenzi etal., Nat Biotechnol, 2021, vol.39, no.11, pp.1453-1465

[0518]

[0038] S. Kishore etal., Genome Biol, 2013, vol.14, no.5, p. R45

[0519]

[0039] S. Kishore etal., Nat Methods, 2011, vol.8, no.7, pp.559-564

[0520]

[0040] A. D. Walsh etal., Commun Biol, 2023, vol.6, no.1, p.555

[0521]

[0041] L. Jin etal., Nat Commun, 2021, vol.12, no.1, p.3715

[0522]

[0042] A. Kozomara etal., Nucleic Acids Res, 2019, vol.47, no. D1, pp. D155-D162

[0523]

[0043] C. Backes etal., Nucleic Acids Res, 2018, vol.46, no. D1, pp. D160-D167

[0524]

[0044] D. H. Lau etal., Cancer Res, 1991, vol.51, no.19, pp.5181-7

[0525]

[0045] J. Suryatenggara et al., Brief Bioinform, 2022, vol.23, no.1, p. bbab537

Claims

CLAIMS1 . Method for in vitro determining a change of a biological state of a subject, comprising the steps(a) providing at least two mirxons from a reference sample, which represent a set of biological states, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), whereinN can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and(CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches,(b) providing the same mirxons of (a) from a sample of the subject,(c) quantifying the levels of one or more of said mirxons and correlating the levels of the mirxons of (a) with the levels of the mirxons of (b), wherein any deviation of said levels reflects differences in said biological states, which are indicative for the change of the biological state to be determined, and(d) normalizing the levels of the mirxons of (c) to the levels of one or more normalizers from the samples, or to a subset of mirxons or all mirxons of (a) or to other RNAs, wherein the levels of said normalizers are most constant over the biological states.

2. The method of claim 1 , wherein the reference sample is from one or more subjects of the same species, specifically said sample is from healthy subjects, subjects of one or more biological states.

3. Method for in vitro monitoring the treatment of a subject, comprising the steps(a) providing at least two mirxons from a reference sample, which represent a set of biological states, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx (SEQ ID NO: 2), whereinN can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and(CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches,(b) providing the same mirxons of (a) from a sample of the subject,(c) quantifying the levels of one or more of said mirxons and correlating the levels of the mirxons of (a) with the levels of the mirxons of (b), wherein any deviation of said levels reflects differences in said biological states, which are indicative for the change of the biological state to be determined, and(d) normalizing the levels of the mirxons of (c) to the levels of one or more normalizers from the samples, or to a subset of mirxons or to all mirxons of (a) or to other RNAs, wherein the levels of said normalizers are most constant over the biological states.

4. The method of any one of claims 1 to 3, wherein the normalized levels are analyzed with a mathematical or logical method to provide information on the probability of a biological state, condition, or change of the subject.

5. The method of claim 4, wherein the normalized levels are combined with further RNA levels, or biological, clinical, environmental, experimental, or biographical data.

6. The method of any one of claims 1 to 5, wherein the normalizers are selected from the group consisting of RNAs, small RNAs, functional RNAs, transfer RNAs or their fragments, circular RNAs, large RNAs, coding RNAs, non-coding RNAs, peptides, proteins, DNA sequences, metabolites, lipids, carbohydrates, internal controls, cell counts, tissue weights, and any combination or subset thereof.

7. The method of any one of claims 1 to 6, wherein the difference in the levels of the mirxons is determined by endpoint, quantitative or digital PCR, miRCURY assays, sequencing, microarrays, Luminex nucleic acid assays, Nanostring assays, or other hybridization-based techniques.

8. The method of any one of claims 1 to 7, wherein the sample is selected from the group of blood, extracellular vesicle, tissue, cell, cell cluster, organoid, spheroid, cell aggregate, ascites, pleural effusion, stool, urine, sperm, saliva, menstrual blood, bile, bone marrow aspirate, breast milk, cerebrospinal fluid, plasma, serum, sweat, sputum, synovial fluid, lymph, oral swab, nasal swap, oropharyngeal swap, nasopharyngeal swap, vaginal swab, urethral swab, cervical swab or lavage, smear of a surface, and body cavity or cell free supernatant of in vitro grown tissues, cells or cell aggregates.

9. Use of the method of any one of claims 1 to 8, for monitoring a subject, specifically for monitoring the development, progression, or change of a biological state.

10. The method of any one of claims 1 to 9, wherein the biological states are selected from the group consisting of biological, physical, chemical, clinical, developmental and pathological states, specifically malignant disease, cardiovascular disease, endocrine disorder and disease, immunological disease, mental disorder, psychological disorder and disease, infection, malnutrition, intoxication, injury, cellular sternness, cellular developmental state, cellular degree of differentiation, or a selection thereof.

11. The method of any one of claims 1 to 10, wherein the mirxons comprise mismatches at the C nucleotides of the (CGG)e (SEQ ID NO: 3) nucleotide sequence.

12. The method of any one of claims 1 to 10, wherein the mirxons comprise mismatches at the G nucleotides of the (CCG)e (SEQ ID NO: 4) nucleotide sequence.

13. The method of any one of claims 1 to 12, wherein the mirxons comprise mismatches which are one-base deletions or insertions.

14. A library comprising ten or more mirxons, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)6Ny(SEQ ID NO: 1) or Ny(CCG)6Nx(SEQ ID NO: 2), whereinN can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and(CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

15. The library of claim 14, wherein said library is constructed by isolating and selecting one or more mirxons from one or more samples.

16. A reference sample library comprising data sets of at at least ten mirxon levels, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)eNy(SEQ ID NO: 1) or Ny(CCG)6Nx(SEQ ID NO: 2), whereinN can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and(CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

17. Composition comprising one or more mirxons, wherein said mirxons are small RNAs consisting of the sequence Nx(CGG)eNy (SEQ ID NO: 1) or Ny(CCG)6Nx(SEQ ID NO: 2), whereinN can be any nucleotide, y and x are non-negative integers, the sum of x and y is not bigger than 6, and(CGG)e (SEQ ID NO: 3) or (CCG)e (SEQ ID NO: 4) can comprise up to 6 nucleotide mismatches.

18. The composition of claim 17, wherein the bases or backbone of said mirxons are replaced by molecules not naturally occurring in RNA molecules, modified ribose sugars, or amide linkages19. The composition of claim 18, wherein said molecules are deoxyribonucleotide bases, artificial bases, locked nucleic acid (LNA) bases, or equivalent molecules.

20. Pharmaceutical composition comprising the composition of claim 13 or 14 and pharmaceutical acceptable excipients.

21. The composition of any one of claims 17 to 19 or the pharmaceutical composition of claim 20 for use in the treatment of a disease or disorder, specifically selected from the group consisting of malignant disease, cardiovascular disease, endocrine disorder and disease, immunological disease, mental disorder, psychological disorder and disease, infection, malnutrition, intoxication, injury.

Images