Method for differentiating patients across healthy control, bipolar disorder, schizophrenia, and schizoaffective groups

Abstract

The present invention relates to a method for differentiating patients across healthy control, bipolar disorder, schizophrenia, and schizoaffective groups. The present invention is drawn to an in vitro method for differential diagnosis of schizophrenia versus schizoaffective versus bipolar disorder in a patient the method comprising the determination in a blood sample of the patient of the relative proportion of at least one of the RNA editing-based biomarkers, IFNAR1, IFNAR2 and AHR, exhibiting A-to-l RNA editing. The present invention also relates to a method for determining the effectiveness of a therapy, or to predict or monitor treatment of such disorder and to a kit to implement the method of the present invention using pairs of primers.

Description

[0001] Method for differentiating patients across healthy control, bipolar disorder, schizophrenia, and schizoaffective groups The present invention relates to a method for differentiating patients across healthy control, bipolar disorder, schizophrenia, and schizoaffective groups. The present invention is drawn to an in vitro method for differential diagnosis of schizophrenia versus schizoaffective versus bipolar disorder in a patient, the method comprising the determination in a blood sample of the patient of the relative proportion of RNA editing-based biomarker (s) of at least one target gene exhibiting A-to-I RNA editing. Mental disorders are highly prevalent, and often devastating diseases, that negatively impact the lives of millions of people worldwide (Scangos et al., 2023). Bipolar Disorder (BD) is a severe, recurrent and often disabling mood disorder characterized by manic, hypomanic or mixed episodes, and alternating episodes of depression (Pacchiarotti et al., 2019). Schizophrenia (SZ) is characterized by positive symptoms such as delusions and hallucinations, and negative symptoms including avolition, social withdrawal, and blunted affect (Nolan et al., 2023). Significantly, schizoaffective (SA) disorder is a persistent and severe illness characterized by the simultaneous manifestation of symptoms associated with schizophrenia (SZ) and affective disorders, including depression and / or mania (Pacchiarotti et al., 2019). The reliable and accurate differential diagnosis presents a notable challenge for the appropriate treatment of these mental disorders (Salvetat et al., 2022). Recently, we identified modifications in RNA editing mediated by ADARs, highlighting that a combination of 8 blood RNA editing- related genes (namely PRKCB, PDE8A, CAMK1D, GAB2, IFNAR1, KCNJ15, LYN, and MDM2) enables the differentiation between patients with BD and unipolar depression (Salvetat et al., 2022). Differential diagnosis of severe mental disorders such as bipolar disorder, schizophrenia, and schizoaffective disorder is still challenging. We previously demonstrated the importance of RNA editing biomarkers for diagnosis in various bipolar disorder cohorts. Thus, the need for a reliable and accurate differential diagnosis, allowing an adequate treatment, of these pathologies has become a crucial necessity for the coming years. Surprisingly, leveraging RNA editing and artificial intelligence analysis, the inventors have successfully differentiated individuals across healthy control, bipolar disorder, schizophrenia, and schizoaffective groups. For example, the inventors have demonstrated that the examination of a panel of RNA editing biomarkers belonging to these 8 genes or the use of IFNAR2 RNA editing-based biomarkers enabled the differentiation of individuals with SZ, SA, BD and healthy control (Ctrl) with high accuracy (see the following examples , tables and figures). This finding confirms the potential of leveraging artificial intelligence (AI)-based predictions using RNA-editing biomarkers. The present in vitro method can patients exhibiting one of said disorders. It can be used not only to support the diagnosis, prognosis but also to choose the most appropriate treatment for these patients. In a first aspect the present invention is directed to a method for in vitro differential diagnosis of schizophrenia versus schizoaffective versus bipolar disorder versus healthy control in a patient, the method comprising the determination in a blood sample of the patient of the relative proportion of RNA editing-based biomarker (s) of at least one target gene exhibiting A-to-I RNA editing. In a preferred embodiment of the method of the present invention,, said at least one target gene exhibiting A-to-I RNA editing is selected from the group of IFNAR1, IFNAR2 and AHR gene The present invention also relates to the use of the relative proportion of RNA editing-based biomarker (s) of at least one target gene exhibiting A-to-I RNA editing in a blood sample of a patient for in vitro differential diagnosis of schizophrenia versus schizoaffective versus bipolar disorder versus healthy control in said patient. In a preferred embodiment of the use of the present invention, said at least one target gene exhibiting A-to-I RNA editing is selected from the group of IFNAR1, IFNAR2 and AHR gene. According to the present invention, said method can be used for differentiating patients across healthy control, bipolar disorder, schizophrenia, and schizoaffective groups In a preferred embodiment, the method of the present invention is a method allowing to differentiate in a patient : - schizophrenia (SCZ) versus healthy control (CTRL); or - SCZ versus bipolar disorder (BP); or - SCZ versus schizoaffective (SCA); or - SCA versus CRTL; or - SCA versus BP or - SCZall (SCZ +SCA) versus CTRL or - SCZall versus BP versus CTRL or - SCZ versus SCA versus SCZall versus BP versus CRTL. In a preferred embodiment, the method of the present invention is a method allowing to differentiate in a patient or to diagnose whether the patient to be tested is suffering from: : - schizophrenia (SCZ) versus healthy control (CTRL; or - SCZ versus bipolar disorder (BP); or - SCZ versus schizoaffective (SCA); or - SCA versus CTRL; or - SCA versus BP or - SCZall (SCZ +SCA) versus CTRL or - SCZall versus BP versus CTRL or - SCZ versus SCA versus SCZall versus BP versus CTRL. In a preferred embodiment, said biomarker is the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said target gene, In a preferred embodiment, said blood sample of the patient is obtained from a patient during depression phase, preferably during moderate or severe depression phase. In a preferred embodiment, said target A-to-I editing RNA gene(s) (mediated by ADARs) is / are: - selected from the group consisting of AHR, CAMK1D, CREB1, FLNB,, GAB2, IFNAR1, IFNAR2, IL17RA, KCNJ15, LYN, MDM2, OTUD7B, PDE8A, PIAS1, PRKCB, PTPRC, RAB36, RASSF1, TMEM63B and ZNF267 gene; preferably - selected from the group consisting of AHR, CAMK1D, FLNB, GAB2, IFNAR1, IFNAR2, IL17RA, KCNJ15, LYN, MDM2, OTUD7B, PDE8A, PIAS1, PRKCB, PTPRC, RAB36, RASSF1 and ZNF267. In a preferred embodiment of the present invention, it is provided a method for in vitro differential diagnosis of SCZ versus SCA versus SCZall versus BP versus CRTL in patient, , said method comprising the step of: a) determining in a sample of the patient to be tested the relative proportion of RNA editing at at least a given editing site and / or isoform and / or pattern, or combination of given sites and / or isoforms and / or patterns, which can be edited on the selected target mRNA gene(s);. b) determining a value resulting from the proportion(s) obtained in step a); c) classifying said patient as being SCZ versus SCA versus SCZall versus BP after comparing said resulting value obtained in step b) to a control value obtained for these disorders, wherein said control values were determined in a manner comparable to that of the resulting value obtained in step b). In a preferred embodiment, in step c) of said methods according to the present invention, said resulting value obtained for the patient to be tested is compared to a threshold associated to the disorder which is desired to be identified (for example, but non-limited to a cut-off or a probability), classifying said patient as being SCZ, SCA, SCZall or BP patient according to the chosen threshold if said resulting value is equal to, or greater or smaller than said threshold. Are also preferred the in vitro method for differential diagnosing according to present invention, wherein in step b), the resulting value is calculated by an algorithm implementing a multivariate method including for example : - mROC program, particularly to identify the linear combination, which maximizes the AUC (Area Under the Curve) ROC and wherein the equation for the respective combination is provided and can be used as a new virtual marker Z, as follows: Z = a.(Biomarker 1) + b.(Biomarker 2) + …i.(Biomarker i) +….n .(Biomarker n) where i are calculated coefficients and (Biomarker i) are the level of the considered biomarker (i.e. level of RNA editing site or isoforms or pattern for a given target / biomarker); and / or - a Random Forest (RF) approach applied to assess the RNA editing site(s) and / or isoforms combinations, particularly to rank the importance of the RNA editing site(s) and / or isoform(s), and to combine the best RNA editing site(s) and / or isoform(s), and / or optionally - a Extra trees (short for extremely randomized trees) approach and all other ensemble- supervised machine learning methods that use decision trees, - a XGboost (short for eXtreme Gradient Boosting) approach and all other gradient boosting algorithm, - a multivariate analysis applied to assess the RNA editing site(s) and / or isoforms combinations for the diagnostic, said multivariate analysis being selecting for example from the group consisting of: - Logistic regression model and penalized logistic regression (as LASSO, ridge or elasticNet methods) applied for univariate and multivariate analysis to estimate the relative risk of patient at different level of RNA editing site and / or isoforms and / or patterns values. - Support Vector Machine (SVM) approach, - Artificial Neural Network (ANN) approach and all other machine learning methods based on neural networks; - Bayesian network approach; - WKNN (weighted k-nearest neighbours) approach; - Any other mathematical method that combines biomarkers.,-and any other method that combine algorithms The resulting value can also be the combination of several models using 1 or more of the algorithms described above, forming a super-learner using, for example, a stacking method.. The sample of the patient to be tested is a biological sample, preferred is a blood, serum, urine, saliva, sweat, tear fluid or plasma sample, more preferably a blood sample, blood serum, even more preferably the sample of the patient is a blood sample, even more preferably a whole blood sample. In a preferred embodiment, the present invention relates to the method or the use according to the present invention wherein the preferred editing sites, isoform and pattern biomarkers according to the selected target A-to-I editing RNA gene and the elected type of diagnostic (CRTL versus SCA, CRTL versus SCZ, CRTL versus SCZall, and SCZ versus SCA) are those indicated in the Tables 8 to 11). The relative proportion of RNA editing at a given editing ‘site’ represents the sum of editing modifications measured at this unique genomic coordinate. Conversely, an edited mRNA isoform is a unique molecule that may or may not contain multiple editing modifications on the same transcript. For example for a given transcript, the edited mRNA isoform BC contains an A-to-I modification on both site B and site C within the same transcript. An editing pattern is one combination of editing events occurring at sites of interest in a transcript. By definition, given a list of sites of interest, a transcript has as many editing patterns there are possible combinations. For example for a given transcript, with sites of interest ABC, the analyzed patterns are [A,B,C,AB,BC,AC, ABC]. The terms “biomarker” refers to an editing site or isoform or pattern of RNA that contains one or more positions differently edited, The method or the use of the present invention wherein: - the diagnosing CRTL versus SCZall, is carried out by using the 18 editing sites, isoform and pattern biomarkers of the selected 7 target A-to-I editing RNA gene indicated in the Table 11 - the diagnosing CRTL versus SCZ, is carried out by using the 10 editing sites, isoform and pattern biomarkers of the selected 7 target A-to-I editing RNA gene indicated in the Table 12 ) or by using the corresponding mROC formula indicated in Table 15; - the diagnosing CRTL versus SCA, is carried out by using the 20 editing sites, isoform and pattern biomarkers of the selected 9 target A-to-I editing RNA gene indicated in the Table 13 or by using the corresponding mROC formula indicated in Table 15 and - the diagnosing SCZ versus SCA, is carried out by using the 14 editing sites, isoform and pattern biomarkers of the selected 5 target A-to-I editing RNA gene indicated in the Table 14 or by using the corresponding mROC formula indicated in Table 15. The method or the use of the present invention wherein at least 2, 3, 4, 5, 6, 7 or 8 target A-to- I editing RNA genes are selected from the group consisting of AHR, CAMK1D, FLNB, GAB2, IFNAR1, IFNAR2, IL17RA, KCNJ15, LYN, MDM2, PDE8A, PIAS1, PRKCB, PTPRC, RAB36, RASSF1 and ZNF267 gene and wherein the biomarkers are the relative proportions of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of these at least selected 2, 3, 4, 5, 6, 7 or 8 target A-to-I editing RNA genes. The method or the use of the present invention wherein the 8 target A-to-I editing RNA genes (targets) CAMK1D, GAB2, IFNAR1, KCNJ15, LYN, MDM2, PDE8A, PRKCB are selected and wherein the biomarkers are the relative proportions of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of these 8 target A-to-I editing RNA genes. The method or the use of the present invention wherein the preferred editing sites, isoform and pattern biomarkers according to the selected diagnostic are those indicated in the Tables 4, 5, 6 and 7. The method or the use of the present invention wherein said target A-to-I editing RNA gene is IFNAR2 gene and the biomarker (s) is / are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of the IFNAR2 gene, preferred are at least 1, more preferred 2, 3, 4, 5 of the biomarkers indicated in Table 20 according to the elected diagnose type. The method or the use of the present invention wherein the 11 target A-to-I editing RNA genes CAMK1D, FLNB, GAB2, IFNAR1, IFNAR2, KCNJ15, PIAS1, PRKCB, PTPRC, RASSF1 and ZNF267 are selected for diagnosing SCZall patients versus CTRL and the biomarkers are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said 11 genes. The method or the use of the present invention wherein the 6 target A-to-I editing RNA genes AHR, FLNB, IFNAR1, IFNAR2, PIAS1 and PTPRC are selected for diagnosing SCZ versus CRTL and the biomarker are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said 6 genes. The method or the use of the present invention wherein the 9 target A-to-I editing RNA genes AHR, CAMK1D, GAB2, IFNAR2, IL17RA, PRKCB, PTPRC, RASSF1 and ZNF267 are selected for diagnosing SCA patient versus CTRL, and the biomarker are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said 9 genes . The method or the use of the present invention wherein the 6 target A-to-I editing RNA genes AHR, GAB2, IFNAR2, PRKCB, PTPRC, RASSF1 are selected for diagnosing SCZ patient versus SCA patient, and the biomarker are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said 6 genes. The method or the use of the present invention wherein the PCR products of the gene region of interest of the selected target A-to-I editing RNA gene(s) are sequenced by NGS. The method or the use of the present invention wherein the pair (s) of primers used for obtaining the PCR product(s) of interest of the target A-to-I editing RNA gene(s) is / are given in Tables 2 and 3. The method or the use of the present invention wherein said method or use comprises the step of : a) determining in a sample of the patient to be tested the relative proportion of RNA editing at at least a given editing site and / or isoform and / or pattern, or combination of given sites and / or isoforms and / or patterns, which can be edited on the selected target A-to-I editing RNA gene(s);. b) determining a value resulting from the relative proportion(s) obtained in step a); c) classifying said patient as being: - SCZ versus CTRL; or - SCZ versus BP; or - SCZ versus SCA; or - SCA versus CTRL; or - SCA versus BP or - SCZall versus CTRL; or - SCZall versus BP versus CTRL; or - SCZ versus SCA versus SCZall versus BP versus CTRL ; or . -SCZ versus SCA versus SCZall versus BP versus CTRL, after comparing said resulting value obtained in step b) to a control value obtained for SCZ, SCA, SCZall, BP and / or CTRL patients, wherein said control values were determined in a manner comparable to that of the resulting value obtained in step b). The method or the use of the present invention wherein in step b) said resulting value obtained for the patient to be tested is compared to a threshold associated to the disorder which is desired to be identified (for example, but non-limited to a cut-off or a probability), classifying said patient as being SCZ, SCA , SCZall, BP or CTRL patient according to the chosen threshold if said resulting value is equal to , or greater or smaller than said threshold. In a second aspect of the present invention, a method for determining the effectiveness of a therapy in a subject exhibiting SCZ, SCA or BP disorder or for predicting or monitoring therapy response in said patient is provided, said method comprising: I-analyzing the expression of the RNA editing-based biomarker according to the method or to the use of claims of the present invention before and / or during and / or after treatment of the patient; and II-using the expression profiles determined in step I for each of the analyses for a combined analysis of the results. In another embodiment, the present invention is directed to a method for monitoring treatment for patient suffering from SCZ, SCA or BP disorder and from a blood sample of said patient, said method comprising: A) differential diagnosing of SCZ, SCA, SCZall or BP disorder in said human by the method or the use of the present invention before the beginning of the treatment which is desired to be monitored, B) repeating steps (a) to c) of the method or use of the present invention after a period of time during which said patient receives the desired treatment for said diagnosed SCZ, SCA, SCZall or BP disorder in order to obtain a post-treatment result value, C) comparing the post-treatment result value from step (c) to the result value obtained before treatment. In a third aspect, the present invention is directed to a kit for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient to be tested, said kit comprising: 1) - optionally, instructions to apply the method or the use of claims 1 to for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient to obtain a resulting value determining whether said patient presents SCZ, SCA, SCZall or BP disorder; and 2) a) the pair of primers depicted in Table 3 for the target IFNAR2 gene having the sequence SEQ ID No.6 (forward primer) and the sequence of its reverse primer b) optionally at least another pair, preferably 2, 3, 4, 5, 6 or the 7 pairs of primers selected from the group of pairs of primers depicted in Table 3. In another embodiment, the present invention is directed to a kit for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient to be tested, said kit comprising: 1) - optionally, instructions to apply the method or the use of claims 1 to for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient, to obtain a resulting value determining whether said patient presents SCZ, SCA, SCZall or BP disorder; and 2) a) a pair of primers allowing to obtain an amplicon including or identical to the amplicon obtainable by the pair of primers having the pair of primers depicted in Table 3 for the target gene IFNAR2, gene having the sequence SEQ ID No.6 (forward primer) and the sequence of its reverse primer and b) optionally at least another pair, preferably 2, 3, 4, 5, 6 or the 7 pairs of primers allowing to obtain amplicon(s) including or identical to the amplicon(s) obtainable by the pairs of primers selecting from the group of pair of primers depicted in Table 3. In another preferred embodiment, the present invention is directed to a kit for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient said kit comprising: 1) - optionally, instructions to apply the method or the use of claims 1 to for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient, to obtain a resulting value determining whether said patient presents SCZ, SCA, SCZall or BP disorder; and 2) a) a pair of primers allowing to obtain an amplicon including or identical to the amplicon obtainable by the pair of primers having the pair of primers depicted in Table 3 for the target AHR having the sequence SEQ ID No.1 and the sequence of its reverse primer and b) optionally at least another pair, preferably 2, 3, 4, 5, 6 or the 7 pairs of primers allowing to obtain amplicon(s) including or identical to the amplicon(s) obtainable by the pairs of primers selecting from the group of pair of primers depicted in Table 3. EXAMPLE 1: Materials and methods 1 Subjects and clinical assessment The patient samples included in this study originated from two distinct cohorts. The Helsinki Declaration's guiding principles were closely followed throughout the study. All participants, aged between 18 and 65 years, signed a written informed consent before entering the study. The cohort of BD patients was approved by the Research Ethics Committee of UNIFESP [CEP No. 1427 / 16]. The Structured Clinical Interview for DSM-IV Axis I disorders (SCID-1) was conducted with eligible participants. The Hamilton Depression Rating Scale-17 items (HDRS) and the Young Mania Rating Scale (YMRS) were used to assess the severity of depressive and manic symptoms respectively. The cohort of SZ and SA patients was approved by Western Institutional review boards (WIRB). Patients in this study were diagnosed with SZ according to M.I.N.I. (Mini International Neuropsychiatric Interview) and SCI-PANSS (Structured Clinical interview for the positive and Negative syndrome scale). A form document completed by the subjects’ psychiatrist confirmed their diagnosis according to DSM-IV criteria. Volunteers in the healthy control group included individuals without a history of psychotropic medication use, no lifetime or current mental disorders, and no family history of a major psychiatric disorder in first degree relatives. More details about the demographic and clinical characteristics are in Table 1. 2 RNA extraction and qualification Samples were retrieved in PAXgene™ blood RNA tubes and extracted using MagNA Pure 96 instrument (Roche). Total RNA concentrations and quality were determined with Qubit Fluorometer (Life technologies) and LabChip GX (Perkin-Elmer) instruments, respectively. 3 Targeted next generation sequencing Validated primers were used to amplify each gene on a Peqstar 96 × thermocycler (VWR). PCR products underwent purification with magnetic beads, quantification and indexing. The resulting library was pooled, purified, denatured, spiked with PhiX Control V3, and loaded onto a sequencing cartridge. Finally, the library was sequenced using Illumina NextSeq 500. 4 Bioinformatics analysis FastQC software was used to verify the quality of the sequencing data. A minimal sequencing depth of 10,000 reads for each sample was considered for further analysis. A pre-treatment step was performed, assessing both length and quality scores. Bowtie2 was used to align the processed reads to the reference human genome sequence, GRCh38 (Langmead and Salzberg, 2012). For SVN calling, SAMtools mpileup was applied (Li, 2011). By counting the different nucleotides in each genomic region, proprietary scripts were used to identify the edited positions in the alignment. Each transcript's percentage of all potential biomarkers (sites, isoforms or patterns) was calculated, and a minimum value of a relative proportion of at least 0.1% was established as described in Methods. 5 Biostatistical analysis All statistics and figures were computed with the "R / Bioconductor" software (Gentleman et al., 2004; team R.d.C, 2010). A “Target Editing Index” (TEI), resuming gene-specific editing values, was calculated by linear combination of significant RNA editing variants maximizing AUC of ROC ( Fig.2A) (Su and Liu, 1993). The TEI was transformed using Boxcox transform aiming to guarantee the normal distribution of data (Box and Cox, 1964). After assessing the target validity with the TEI analysis, we explored a signature of combined biomarkers to classify all classes. Due to numerous significant biomarkers (p-value < 0.05), a feature selection process was implemented (caret and FactoMineR R package Kassambara and Mundt, 2020; Kursa and Rudnicki, 2010). A final list of 32 biomarkers belonging to our 8 targets (Hayashi et al., 2023) were combined using the mROC method for each comparison (Fig.1A and 1B). Finally, the same 32 biomarkers were combined using multiclass Random Forest (RF) (Liaw, 2002), considering patient treatment and sex (Table 1). RF requires the use of a training set to construct the model (70% of the population; n=120 samples) and a test set (30% of the population; n=49 samples) to validate it. 6 Results 6.1 Characteristics of the populations This study conducted a comparative analysis involving 85 Ctrl subjects, 39 BD, 31 SZ, and 14 SA patients. Regarding the treatments, five main categories of psychiatric drugs were considered (Table1): antipsychotics, antidepressants, anxiolytics, antiepileptics, and hypnotics / sedatives. Additional details are available in Supplementary Materials. 6.2 Target Editing Index (TEI) An analysis was performed to examine the RNA editing modifications of 8 genes in the 4 groups, summarized as TEI (Fig.2). In comparison to the Ctrl group, the SZ and SA groups showed a significant difference (p-value FDR < 0.10) in TEI on KCNJ15, PRKCB, KCNJ15, and LYN respectively. As compared to SZ and SA groups, the BD group revealed a significant difference (p-value FDR< 0.10) in TEI on GAB2, IFNAR1, KCNJ15, LYN, and MDM2. However, no significant difference in TEI values were noticed between the SZ and SA groups. 6.3 RNA editing biomarkers combination analysis Significant editing biomarkers were selected and combined to identify the best signature for each comparison, facilitating the separation of groups. “Z” represents this combination, comprising a list of 32 significant RNA editing biomarkers. As shown in Fig.1A, 1B and 1C, the combination of (Z1), (Z2), (Z3), (Z4), (Z5) and (Z6) resulted in AUC values of 0.910, 0.955, 0.990, 0.962, 0.999 and 1.00, with sensitivities of 82.05, 91.11, 97.78, 96.77, 100 and 100%, and specificities of 87.06, 87.06, 94.87, 87.06, 98.82, and 100% respectively. This indicates a clear separation between Ctrl and BD, Ctrl and SZ+SA, BD and SZ+SA, Ctrl and SZ, Ctrl and SA, and SZ and SA. 6.4 AI algorithm The 32 significant biomarkers were combined using a multiclass RF algorithm. The algorithm was trained on 70% of the population. Then, the test was made on the 30% of the population who never saw the algorithm. The results obtained were plotted in a 3D scatterplot (see Figure 2A). We separate Ctrl group from individuals with BD, SZ and SA with high AUCs and sensitivities and specificities (see Figure 2B). 7 Discussion Our previous report identified blood RNA editing biomarkers associated to 8 genes holding promise for the diagnosis and treatment of mental disorders (Salvetat et al., 2022). Additionally, these genes could differentiate euthymic from depressed or mixed states BD patients (Hayashi et al., 2023). In this study, we demonstrated that these genes could differentiate individuals with bipolar disorder (BD) from those with schizophrenia (SZ) and schizoaffective (SA) disorder. For this, we selected a list of 32 significant RNA editing biomarkers used for all comparisons (Z1 to Z6, Figures 1A-1C).1), which led to characterize Ctrl from patients with BD, SZ and SA (see Figure 1A-1C).1). Using an RF algorithm, the predictive capacity of these biomarkers to differentiate Ctrl group from patients with BD, SA and SZ was tested (Figure 2B). The 8 biomarkers are expressed in the CNS and most are linked to SZ (CAMK1D, MDM2)(Gadelha et al., 2016; Andrews et al., 2017), IFNAR1 (Singh et al., 2020), LYN, PRKCB (Napolitano et al., 2006), PDE8A (Chimienti et al., 2019) and / or BD (GAB2, KCNJ15) (Salvetat et al, 2022). Some limitations should be taken into account: first, patients in the BD group were in different mood states (hypomanic / manic, depressive, mixed, and euthymic bipolar). Second, sex proportion is significantly different among the groups. Third, limited number of patients presents a potential bias in this study. Increasing the population size, to enhance algorithm’s performance to distinct SZ and SA patients, can potentially improve the present analysis. Nevertheless, this initial proof-of-concept analysis presents compelling evidence for the establishment of an RNA editing signature for diagnosis, and potentially for prognosis and / or treatment prediction. Further validation will be undertaken using a larger cohort. EXAMPLE 2: Supplementary Materials 1. Supplemental Methods 1.1 Bioinformatics analysis of targeted sequencing data The sequencing data obtained from the Illumina NextSeq 500 underwent quality assessment using the FastQC software (version 0.11.7, https: / / github.com / s-andrews / FastQC / ). A minimum sequencing depth of 10,000 reads for each sample and each target was established for further analysis. Initial processing involved removing adapter sequences and filtering based on length and quality score, with short reads (<100 nts) and reads with an average QC < 8 being eliminated. Flexible read trimming and filtering tools for Illumina NGS data, including fastx_toolkit v0.0.14 and prinseq version 0.20.4, were employed to enhance sequence alignment quality. After pre-processing steps, each cleaned fastq file underwent additional quality control before proceeding to further analysis. The processed reads were aligned using bowtie2 (Langmead and Salzberg, 2012)(version 2.2.9) with end-to-end sensitive mode. The alignment was done to the reference human genome sequence GRCh38. Non-unique alignments, unaligned reads, or reads containing insertion / deletion (INDEL) were removed from downstream analysis using SAMtools software (Li et al., 2009)(version 1.7). SAMtools mpileup (Li et al., 2009) was used for SVN calling. Identification of edited positions in the alignment was performed using in-house scripts to determine the number of different nucleotides at each genomic location. For each position, the script calculated the percentage of reads containing a ‘G’ nucleotide [Number of ‘G’ reads / (Number of ‘G’ reads + Number of ‘A’ reads)*100]. Genomic locations with a reference ‘A’ and a percentage of ‘G’ reads > 0.1 were automatically identified by the script as ‘A-to-I edition site’. The final step involves calculating the percentage of all possible isoforms of each transcript. By definition, the relative proportion of RNA editing at a given editing ‘site’ represents the sum of editing modifications measured at this unique genomic coordinate. Conversely, an edited mRNA isoform is a unique molecule that may or may not contain multiple editing modifications on the same transcript. For example, for a given transcript, the edited mRNA isoform BC contains an A-to-I modification on both site B and site C within the same transcript. An editing pattern is one combination of editing events occurring at sites of interest in a transcript. Given a list of sites of interest, a transcript has as many editing patterns as there are possible combinations. For example, for a given transcript, with sites of interest ABC, the analyzed patterns are [A, B, C, AB, BC, AC, and ABC]. We take into consideration the term “biomarker” to refer to an RNA editing site, isoform and or pattern with a significant diagnostic value in at least one studied comparison. 1.2 Biostatistical analysis The software "R / Bioconductor" was used to compute all statistics and figures (Gentleman et al., 2004; team R.d.C, 2010). Resuming gene-specific editing values, a “Target Editing Index” (TEI) was calculated by linearly combining significant RNA editing variants in a way that maximizes the AUC ROC (Salvetat et al., 2022; Su and Liu, 1993). With the goal of ensuring that the data would have a normal distribution, the TEI was transformed using the Boxcox transform (Box and Cox, 1964). A differential analysis was carried out using the most appropriate test between the Mann-Whitney rank-sum test, Student’s t-test or Welch's t-test according to normality and sample variance distribution in each cohort. The false discovery rate (FDR) was managed through the use of the Benjamini and Hochberg (BH) procedure (Benjamini and Hochberg, 1995) and an adjusted p-value below 0.1 was considered as statistically significant. Additionally, all biomarkers were combined for a studied combination in order to evaluate any potential increases in specificity and sensitivity using mROC multivariate approaches (Kramar et al., 2001). mROC determined the linear combination, which maximizes the AUC (Area Under the Curve) ROC (Su and Liu, 1993). The equation for the respective combination is provided and can be used as a new virtual marker Z, as follows: Z = a x biomarker1 + b x biomarker 2 + c x biomarker3, where a, b, c are calculated coefficients and biomarkers 1,2,3 is the level of biomarker. Moreover, all biomarkers were combined with each other to evaluate the potential increase in sensibility and specificity using multiclass random forest (RF), a machine learning approach (Breiman, 2001). This method, plotted in a 3D scatterplot (Fig.2A), was developed with a 70% training population (n=120) and a 30% (n=49) testing population. This sharing has been randomized and respects the initial proportion of the various statutes in each set. RF method combines Breiman’s “bagging” idea and the random selection of features to construct a collection of decision trees with controlled variance. To generalize the model, we performed 1000 learning RF trees. The final classifier was generated by summing up the votes (probabilities) for each applied sample from each tree and was normalized by the number of trees. The end resulting probabilities reflect the majority vote of the 1000 RF trees. We used a grid learning approach for each individual tree, where we stated certain maximum parameter sets (ntree = 1000, nodesize = 1 and mtry = (1,20)). RF results are shown on the test dataset which has never seen the algorithm. The implementation was done using the R randomForest package (version 4.6-14) and R caret package (version 6.0-84) (Kuhn, 2008). 1.3 Characteristics of the populations In this study, 169 participants are included, comprising 85 healthy volunteers, 31 diagnosed with schizophrenia, 14 with schizoaffective disorder, and 39 with bipolar disorder. The t-test was used to compare the age across different groups, while the chi-squared test was employed to compare sex and treatments among the various groups. No significant difference (p > 0.05) was found in the age among the different groups, except between the SA and BD groups compared to the controls (p < 0.05). There was a significant difference (p < 0.05) in sex among all groups except between the schizoaffective and schizophrenic groups. Lastly, a significant difference (p < 0.001) in psychiatric treatment was observed between the BD group and the SA and SZ groups (Table 1), while no significant difference (p > 0.05) was observed between the SA and SZ groups. To prevent a putative bias in further analyses due to differences in medications, the algorithm results were adjusted for sex and psychiatric treatments. Abbreviations of HGNC (HUGO Gene Nomenclature Committee) gene names: ADAR: Adenosine deaminases acting on RNA GAB2: GRB2 Associated Binding Protein 2 IFNAR1: Interferon Alpha and Beta Receptor Subunit 1 LYN: LYN Proto-Oncogene, Src Family Tyrosine Kinase MDM2: MDM2 proto-oncogene PRKCB: Protein Kinase C Beta PDE8A: Phosphodiesterase 8A CAMK1D: Calcium / Calmodulin Dependent Protein Kinase 1D KCNJ15: Potassium inwardly rectifying channel subfamily J member 15 Table 1: Demographic and clinical characteristics of study population. Table 1 Data are presented as the mean ± SEM. P-values for age were obtained using the student’s t- test, while p-values for sex and psychotropic treatments were calculated using Chi-square test. Controls: Healthy volunteers; SZ: Schizophrenia patients; SA: Schizoaffective patients; BD: Bipolar patients EXAMPLE 3: Diagnostic performance of the RF model discriminating Ctrl, Schizophrenia, Schizoaffective and Bipolar Disorders See Figures 1A, 1B and 1C Legend Figures 1A, 1B and 1C. RNA editing signatures to discriminate CTRL vs. BD vs. SA vs. SZ. (1A) Scatterplot 3D of the best combination to discriminate CTRL vs. BD vs. SZ + SA subgroups. (1B) Scatterplot 3D of the best combination to discriminate CTRL vs. SA vs. SZ subgroups. (1C) Table showing per- group comparison performances and decision threshold (cut-off). Z represents a combination of 32 significant RNA editing biomarkers. CTRL: Control; BD: Bipolar disorder; SA: Schizoaffective disorder; SZ: Schizophrenia; AUC: AUC ROC; Sp: Specificity; Se: Sensitivity; PPV: Positive predictive value; NPV: Negative predictive value. See Figures 2A-2B: The AI algorithm was developed by randomly splitting the data into 2 distinct datasets: 70% for training the model (n=120) and 30% (n=49) for testing the model. RF multiclass model was built by optimizing hyper-parameters through cross validation. The model was applied to the test data to assess its performance for internal validation. Legend of Figures 2A-2B: A / Scatterplot 3D of the RF model on the test dataset for discriminate CTRL vs BD vs SA vs SZ subgroups. B / Associated per-class performance metrics of the multiclass RF model (precision, recall and F-1 score), and their macro and weighted averages. Ctrl: Control; BD: Bipolar Disorder; SA: Schizoaffective disorder; SZ: Schizophrenia. EXAMPLE 4: TEI analysis differentiates between the 4 groups (Ctrl, BD, SZ, and SA) See Figures 3A-3B Legend of Figures 3A-3B A / TEI values comparing Ctrl, SZ, SA, and BD; (p-value correspondance: ° : < 0.10, *: <0.05, **: <0.01, ***: < 0.001). B / Differential TEI analysis per-group comparison, showing FDR-adjusted p-values (bold values are < 0.1). Ctrl: Control; BD: Bipolar Disorder; SA: Schizoaffective disorder; SZ: Schizophrenia. EXAMPLE 5: Primers Design, Targeted Sequences See: Table 2 : Primers Table, extended version, and Table 3 : Primers Table, reduced version (only those targets showing significativity with TEI):

[0002] 19 Table 2 : Primers Table, extended version 5 In Table 2, for each of the Primer Forward Sequences SEQ ID N0.1 to 21, the corresponding Primer Reverse Sequences have respectively the SEQ ID N0.22 to 42. Table 3: Primers Table, reduced version (only those targets showing significativity with TEI):

[0003] In Table 3, the numbering of the SEQ IDs (1-16) should not be considered. Only the numbering of the sequences in Table 2 should be considered for the sequence numbering in the sequence listing and / or in the claims. 5 EXAMPLE 6: Significant Biomarkers See: Table 4: Top10 best Biomarkers by Type: CTRL vs SCZ+SCA,Table 5: Top10 best Biomarkers by Type: CTRL vs SCZ, Table 6: Top10 best Biomarkers by Type: CTRL vs SCA, and 10 Table 7: Top10 best Biomarkers by Type: SCZ vs SCA.

[0004] 22 Table 4: Top10 best Biomarkers by Type : CTR L vs SCZ+SCA 23 Table 5: Top10 best Biomarkers by Type : CTRL vs SCZ 24 Table 6: Top10 best Biomarkers by Type : CTRL vs SCA 25 Table 7: Top10 best Biomarkers by Type : SCZ vs SCA EXAMPLE 7: Best 5 significant Biomarkers by Target and Comparison See: 5Table 8: Top5 best Biomarkers by Target: AHR, CAMK1D, FLNB, GAB2, IFNAR1 andIFNAR2, Table 9: Top5 best Bi omarkers by Target: IL17RA, KCNJ15, LYN, MDM2, OTUD7B and PDE8A, and Table 10: Top5 best Biomarkers by Target: PIAS1, PRKCB, PTPRC, RAB36, RASSF1 and 10 ZNF267. Table 8: Top5 best Biomarkers by Target : AHR, CAMK1D, FLNB, GAB2, IFNAR1 and IFNAR2 5 Table 9: Top5 best Biomarkers by Target : IL17RA, KCNJ15, LYN, MDM2, OTUD7B and PDE8A

[0005] Table 10: Top5 best Biomarkers by Target : PIAS1, PRKCB, PTPRC, RAB36, RASSF1 and ZNF267 EXAMPLE 8: Performances mROC par Comparison See: Figures 4A and 4B, and Table 11: mROC. CTRL (n = 44) vs SCZall (SCZ + SCA) (n = 45), Figures 5A and 5B, a n d T ab le 12: m RO C. C TR L (n = 44) vs SCZ (n = 31), Figures 6A and 6B, and Table 13: mROC . CTRL (n = 44) vs SCA (n = 14), Figures 7A and 7B, and Table 1 4: mROC . SCZ (n = 31) vs SCA (n = 14), and e 15: mROC formu la for CTR Tabl L (n = 4 4) vs S CZ (n =31), mR OC. CTRL (n =44) vs SCA (n = 14) and m RO C. SCZ (n = 31) v s SCA (n = 14). Table 11: mROC. CTRL (n = 44 ) vs SCZall (SCZ + SCA) (n = 45) 18 BMKs, 7 targets (IF NAR1, IFNAR2, IL17RA, LYN, PTPRC, R ASSF1, Z NF267)

[0006] 30 Table 12: mROC. CTRL (n = 44) vs SCZ (n = 31) 10 BMKs, 7 targets (CAMK1D, GAB2, IFNAR1, LYN, PIAS1, PRKCB, RASSF1 Table13: mROC. C TRL (n = 44) vs SCA (n = 14) 20 BMKs, 9 targets (AHR, CAMK1D, IFNAR1, I FN AR2, IL17RA, LYN, MDM2, PRKCB, RAB36)

[0007] 31 Table 14: mROC. SCZ (n = 31) vs SCA (n =14)14 BMKs, 5 targets (FLNB, IFNAR2, PRKCB, PTPRC, RASSF1) 5 Table15: mROC formula EXAMPLE 9: 3D Plot using mROC combiZ See: Figure 8: mROC: CTRL-vs-SCZ vs CTRL-vs-SCA vs SCZ-vs-SCA EXAMPLE 10: Target Editing Index (TEI) See:Table 16: Target Editing Index: CTRL (n = 44) vs SCZall (n = 45),Table 17: Target Editing Index: CTRL (n = 44) vs SCZ (n = 31), Table 18: Target Editing Index: CTRL (n = 44) vs SCA (n = 14), Table 19: Target Editing Index: SCZ (n = 31) vs SCA (n = 14), and Figure 9: Target Editing Index Plot, Figures 10A, 10B, 10C and 10D: - 10A: BMKS 11 Targets (CAMK1D, FLNB, GAB2, IFNAR1, IFNAR2, KCNJ15, PIAS1, PRKCB, PTPRC, RASSF1 and ZNF267); -10B: BMKS 6 Targets (AHR, FLNB, IFNAR1, IFNAR2, PIAS1 and PTPRC) -10C: BMKS 9 Targets (AHR, CAMK1D, GAB2, IFNAR2, IL17RA, PRKCB and PTPRC) -10D: BMKS 6 Targets (AHR, GAB2, IFNAR2, PRKCB, PTPRC, RASSF1 Figure 11: When combining together the more significant biomarkers used for previous TEI, we can separate Controls, SCZ and Schizoaffective (figure 11 at left).

[0008] 33 T ab le16: Targ e t Editin g Index : CTRL (n = 44) vs SCZal l (n = 4 5 )

[0009] 34 T abl e 17: Targ et Editin g Index : CTRL (n = 44) vs SCZ ( n = 31) 5 Table 18: Target Editing Index: CTRL (n = 44) vs SCA (n = 14) Table 19: Target Editing Index: SCZ (n = 31) vs SCA (n = 14) 5 EXAMPLE 11: Analysis SCZ vs BD with IFNAR2 only See: Table 20.

[0010] Table 20: Top30 best Biomarkers with AUC > 0.7 BMQnames CTRL_vs_BD CTRL_vs_SCZ CTRL_vs_SCA BD_vs_SCZ BD_vs_SCA SCZ_vs_SCA IFNAR2_iso_B NA NA NA NA p-value = 1e-04 ; AUC = 0.904 NA ; ; ; ; Figures 20A and 20B and Table 21: RNA editing signatures for discriminate CTRL vs BD vs SA vs SZ. -Figure 20A: Scatterplot 3D of the best combination to discriminate CTRL vs BD vs SZ+SA subgroups -Figur e 20B: Sc atter plo t 3D of th e b est com bi na ti on to discr im inat e C TR L v s S A vs SZ subgroups. Table 21: Table showin g per-group comparison performances and decision thr esh old (cut-o ff). Z represents a combination of 32 significant RNA editing biomarkers. Ctrl: Control; BD: Bipolar Disorder; SA: Schizoaffective disorder; SZ: Schizophrenia, AUC=AUC ROC; Sp=Specificity; Se=Sensitivity; PPV= Positive Predictive Value; NPV= Negative Predictive Value. Table 21: Figures 21A and Table22: TEI analysis differentiates between the 4 groups (Ctrl, BD, SZ, and SA) - Figure 21: TEI values comparing Ctrl, SZ, SA, and BD ; (p-value correspondance: « ° » : < 0.10, « * » : <0.05, « ** » : <0.01, « *** » : < 0.001). - Table 22: Differential TEI analysis per-group comparison, showing FDR-adjusted p- values (bold values are < 0.1). 38 Table 22 Table 23, Figures 12A and12B: mROC. CTRL (n = 74) vs BD (2BSB cohort) (n = 29),Table 24, Figures 13A and 13B: mROC. CTRL (n = 74) vs SCZ + SCA (n = 45), Table 25, Figures 14A and 14B: mROC. CTRL (n = 74) vs SCZ (n = 31), Table 26, Figures 15 A and 1 5B: m RO C. CT RL ( n = 74) vs SCAf fective (n = 14), Table 27, Figures 16 A and 16B: mROC. BD (n = 29) vs SCZall (No affe ctive + Af fective) (n = 45), Table 28, Figures 17A and 17 B: mROC . BD (n = 29) vs SCZ (n = 31), Table 29, Figures 18A and 18B: mROC. BD (n = 2 9) vs SCAffective (n = 14), Table 30, Figures 19A and 19B: mROC. SCZ (n = 31) vs SCAffective (n = 14), Table 23: mROC. CTRL (n = 74) vs BD (2BSB cohort) (n = 29) Table 24: mROC. CTRL (n = 74) vs SCZ + SCA (n = 45) 5 Table 25: mROC. CTR L (n =7 4) vs S CZ (n = 31) Table 26: mROC. CTRL (n = 74) vs SCAffective (n = 14)

[0011] 41 Table 27: mROC. BD (n = 29) vs SCZall (No affec tiv e + A ff ective) (n = 45)

[0012] 42 Table 28: mROC. BD (n = 29) vs SCZ (n = 31) 5 Table 29: mROC. BD (n =2 9) vs SCA ffective (n =1 4) Table30: mROC. SCZ (n =31) vs SCAffective (n = 14) 10

[0013] Bibliography: Andrews, J.L et al.,., 2017. Alterations of ubiquitin related proteins in the pathology and development of schizophrenia: Evidence from human and animal studies. J Psychiatr Res 90, 31-39. doi: 10.1016 / j.jpsychires.2017.01.009 Benjamini, Y et al.,, 1995. Controlling The False Discovery Rate - A Practical And Powerful Approach To Multiple Testing. J R. Stat. Soc Ser. B 57, 289–300. https: / / doi.org / 10.2307 / 2346101 B Chimienti, F., Cavarec, L., Vincent, L., Salvetat, N., Arango, V., Underwood, M.D., Mann, J.J., Pujol, J.-F., Weissmann, D., 2019. Brain region-specific alterations of RNA editing in PDE8A mRNA in suicide decedents. Transl Psychiatry 9, 91. doi: 10.1038 / s41398-018-0331-3 FactoMineR : analyse de données avec R [WWW Document], n.d. URL http: / / factominer.free.fr / index_fr.html (accessed 12.22.23). Gadelha, A., et al., 2016. Genome-wide investigation of schizophrenia associated plasma Ndel1 enzyme activity. Schizophr Res 172, 60-67. doi: 10.1016 / j.schres.2016.01.043 Gentleman, R.C et al., 2004. Bioconductor: open software development for computational biology and bioinformatics. Genome Biology 5(10), R80. doi: 10.1186 / gb-2004-5-10-r80. Hayashi, M.A.F., Salvetat, N., Cayzac, C., Checa-Robles, F.J., Dubuc, B., Mereuze, S., Nani, J.V., Molina, F., Brietzke, E., Weissmann, D., 2023. Euthymic and depressed bipolar patients are characterized by different RNA editing patterns in blood. Psychiatry Res 328, 115422. doi: 10.1016 / j.psychres.2023.115422 Kassambara, A et al., 2020. factoextra: Extract and Visualize the Results of Multivariate Data Analyses. R Package Version 1.0.7. https: / / CRAN.R-project.org / package=factoextra Kramar, A et al., 2001. mROC: a computer program for combining tumour markers in predicting disease states. Comput. Methods Programs Biomed. 66, 199–207. https: / / doi.org / 10.1016 / S0169-2607(00)00129-2 Kursa, M.B et al., 2010. Feature Selection with the Boruta Package. Journal of Statistical Software 36, 1-13. doi: 10.18637 / jss.v036.i11 Kuhn, M., 2008. Building Predictive Models in R Using the caret Package. J. Stat. Softw.28. https: / / doi.org / 10.18637 / jss.v028.i05 read alignment with Bowtie 2. Nat. Methods 9, 357–359. https: / / doi.org / 10.1038 / nmeth.1923 Li, H., Handsaker, et al.., 1000 Genome Project Data Processing Subgroup, 2009. The Sequence Alignment / Map format and SAMtools. Bioinformatics 25, 2078–2079. https: / / doi.org / 10.1093 / bioinformatics / btp352 calling, mutation discovery, association mapping and population genetical parameter estimation from sequencing data. Bioinformatics 27, 2987- 2993. doi: 10.1093 / bioinformatics / btr509 Liaw, A., 2002. Classification and Regression by randomForest. R news. Napolitano, M., Picconi, B., Centonze, D., Bernardi, G., Calabresi, P., Gulino, A., 2006. L- DOPA treatment of parkinsonian rats changes the expression of Src, Lyn and PKC kinases. Neurosci Lett 398, 211-214. doi: 10.1016 / j.neulet.2005.12.077 Nolan, C.J et al., 2023. Beyond the Binary: Gender Inclusivity in Schizophrenia Research. Biological Psychiatry, Circuit Dysregulation in Schizophrenia 94, 543-549. doi: 10.1016 / j.biopsych.2023.03.018 Pacchiarotti, I. et al., 2019. Long-acting injectable antipsychotics (LAIs) for maintenance treatment of bipolar and schizoaffective disorders: A systematic review. Eur Neuropsychopharmacol 29, 457-470. doi: 10.1016 / j.euroneuro.2019.02.003 Salvetat, N., Checa-Robles, F.J., Patel, V., Cayzac, C., Dubuc, B., Chimienti, F., Abraham, J.- D., Dupré, P., Vetter, D., Méreuze, S., Lang, J.-P., Kupfer, D.J., Courtet, P., Weissmann, D., 2022. A game changer for bipolar disorder diagnosis using RNA editing-based biomarkers. Transl Psychiatry 12, 182. doi: 10.1038 / s41398-022-01938-6. Scangos, K.W.,et al., 2023. New and emerging approaches to treat psychiatric disorders. Nat Med 29, 317-333. doi: 10.1038 / s41591-022-02197-0. Singh, H et al., 2020. A pivotal role for Interferon-α receptor-1 in neuronal injury induced by HIV-1. J Neuroinflammation 17, 226. doi: 10.1186 / s12974-020-01894-2. Su, J.Q., Liu, J.S., 1993. Linear Combinations of Multiple Diagnostic Markers. Journal of the American Statistical Association 88, 1350-1355. doi: 10.1080 / 01621459.1993.10476417. team R.d.C, 2010. R a language and environment for statistical computing: reference index. R Foundation for Statistical Computing, Vienna.

Claims

1. CLAIMS 1. A method for in vitro differential diagnosis of schizophrenia versus schizoaffective versus bipolar disorder versus healthy control in a patient, the method comprising the determination in a blood sample of the patient of the relative proportion of RNA editing-based biomarker (s) of at least one target gene exhibiting A-to-I RNA editing, and wherein said at least one target gene exhibiting A-to-I RNA editing is selected from the group of IFNAR1, IFNAR2 and AHR gene 2. The use of the proportion of RNA editing-based biomarker (s) of at least one target gene exhibiting A-to-I RNA editing in a blood sample of a patient for in vitro differential diagnosis of schizophrenia versus schizoaffective versus bipolar disorder versus healthy control in said patient, and wherein said at least one target gene exhibiting A-to-I RNA editing is selected from the group of IFNAR1, IFNAR2 and AHR gene.

3. The method of claim 1 or the use of claim 2, for differentiating patients across healthy control, bipolar disorder, schizophrenia, and schizoaffective groups 4. The method of claim 1 or 3, or the use of claim 2 or 3, wherein the patient is diagnosed: - schizophrenia (SCZ) versus healthy control (CTRL); or - SCZ versus bipolar disorder (BP); or - SCZ versus schizoaffective (SCA); or - SCA versus CTRL; or - SCA versus BP or - SCZall (SCZ +SCA) versus CTRL; or - SCZall versus BP versus CTRL; or - SCZ versus SCA versus SCZall versus BP versus CTRL.

5. The method of claims 1, 3 and 4, or the use of claims 2 to 4, wherein said biomarker is the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said target gene.

6. The method of claims 1 and 3 to 5, or the use of claims 2 to 5 wherein said blood sample is obtained from a patient during depression phase, preferably during moderate or severe depression phase.

7. The method of claims 1 and 3 to 6 or the use of claims 2 to 6 wherein the preferred editing sites, isoform and pattern biomarkers according to the selected target A-to-I editing RNA gene and the elected type of diagnostic (CTRL versus SCA, CTRL versus SCZ, CTRL versus SCZall, and SCZ versus SCA) are those indicated in the Tables 8 to 11.

8. The method of claims 1 and 3 to 7 or the use of claims 2 to 7 wherein: - the diagnosing CTRL versus SCZall, is carried out by using the 18 editing sites, isoform and pattern biomarkers of the selected 7 target A-to-I editing RNA gene indicated in the Table 11; - the diagnosing CTRL versus SCZ, is carried out by using the 10 editing sites, isoform and pattern biomarkers of the selected 7 target A-to-I editing RNA gene indicated in the Table 12 or by using the corresponding mROC formula indicated in Table 15; - the diagnosing CTRL versus SCA, is carried out by using the 20 editing sites, isoform and pattern biomarkers of the selected 9 target A-to-I editing RNA gene indicated in the Table 13 or by using the corresponding mROC formula indicated in Table 15 ; and - the diagnosing SCZ versus SCA, is carried out by using the 14 editing sites, isoform and pattern biomarkers of the selected 5 target A-to-I editing RNA gene indicated in the Table 14 or by using the corresponding mROC formula indicated in Table 15.

9. The method of claims 1 and 3 to 8 or the use of claims 2 to 8 wherein 8 target A-to-I editing RNA genes CAMK1D, GAB2, IFNAR1, KCNJ15, LYN, MDM2, PDE8A, PRKCB are selected and wherein the biomarkers are the relative proportions of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of these 8 target A-to-I editing RNA genes.

10. The method of claims 1 and 3 to 9 or the use of claims 2 to 9 wherein the preferred editing sites, isoform and pattern biomarkers according to the elected diagnostic are those indicated in the Tables 4, 5, 6 and 7.

11. The method of claims 1 and 3 to or the use of claims 2 to 7 wherein said target A-to-I editing RNA gene is IFNAR2 gene and the biomarker (s) is / are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of the IFNAR2 gene, preferred are at least 1, more preferred 2, 3, 4, 5 of the biomarkers indicated in Table 20 according to the elected diagnose type.

12. The method of claims 1, 3 to 12, or the use of claims 2 to 11, wherein 11 target A-to-I editing RNA genes CAMK1D, FLNB, GAB2, IFNAR1, IFNAR2, KCNJ15, PIAS1, PRKCB, PTPRC, RASSF1 and ZNF267 are selected for diagnosing SCZall patients versus CTRL and the biomarkers are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said 11 genes.

13. The method of claims 1, 3 to 12, or the use of claims 2 to 12 wherein 6 target A-to-I editing RNA genes AHR, FLNB, IFNAR1, IFNAR2, PIAS1 and PTPRC are selected for diagnosing SCZ versus CTRL and the biomarker are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said 6 genes.

14. The method of claims 1, 3 to 13, or the use of claims 2 to 13 wherein 9 target A-to-I editing RNA genes AHR, CAMK1D, GAB2, IFNAR2, IL17RA, PRKCB, PTPRC, RASSF1 and ZNF267 are selected for diagnosing SCA patient versus CTRL, and the biomarker are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said 9 genes.

15. The method of claims 1, 3 to 14, or the use of claims 2 to 14, wherein 6 target A-to-I editing RNA genes AHR, GAB2, IFNAR2, PRKCB, PTPRC, RASSF1 are selected for diagnosing SCZ patient versus SCA patient, and the biomarker are the relative proportion of RNA editing at a given editing site and / or isoform and / or pattern or a combination of given sites and / or isoforms and / or patterns of said 6 genes.

16. The method of one of claims 1 to 15 or the use of claims 2 to 15, wherein the pair (s) of primers used for obtaining the PCR product(s) of interest of the target A-to-I editing RNA gene(s) is / are given in Tables 2 and 3:

17. The method of claims 1 and 3 to 16, or the use of claims 2 to 16 the method comprising the step of : a) determining in a sample of the patient to be tested the relative proportion of RNA editing at at least a given editing site and / or isoform and / or pattern, or combination of given sites and / or isoforms and / or patterns, which can be edited on the selected target A-to-I editing RNA gene(s);. b) determining a value resulting from the proportion(s) obtained in step a); c) classifying said patient as being: - SCZ versus CTRL; or - SCZ versus BP; or - SCZ versus SCA; or - SCA versus CTRL; or - SCA versus BP or - SCZall versus CTRL; or - SCZall versus BP versus CTRL; or - SCZ versus SCA versus SCZall versus BP versus CTRL ; or . -SCZ versus SCA versus SCZall versus BP versus CTRL, after comparing said resulting value obtained in step b) to a control value obtained for SCZ, SCA , SCZall, BP and / or CTRL patients, wherein said control values were determined in a manner comparable to that of the resulting value obtained in step b).

18. The method or the use of claim 17, wherein in step b) said resulting value obtained for the patient to be tested is compared to a threshold associated to the disorder which is desired to be identified (for example, but non-limited to a cut-off or a probability), classifying said patient as being SCZ, SCA , SCZall, BP or CTRL patient according to the chosen threshold if said resulting value is equal to , or greater or smaller than said threshold.

19. A method for determining the effectiveness of a therapy in a subject exhibiting SCZ, SCA or BP disorder or for predicting or monitoring therapy response in said patient is provided, comprising:I-Analyzing the expression of the RNA editing-based biomarker according to the method or to the use of claims 1 to 18 before and / or during and / or after treatment of the patient; and II-Using the expression profiles determined in step I for each of the analyses for a combined analysis of the results.

20. A method for monitoring treatment for SCZ, SCA or BP disorder, from a blood sample of said patient, said method comprising: A) differential diagnosing of SCZ, SCA, SCZall or BP disorder in said human by the method or the use of claims 1 to 18 before the beginning of the treatment which is desired to be monitored; B) repeating steps (a) to c) of the method or use of claims 1 to 18 after a period of time during which said patient receives the desired treatment for said diagnosed SCZ, SCA, SCZall or BP disorder in order to obtain a post-treatment result value; C) comparing the post-treatment result value from step (c) to the result value obtained before treatment.

21. A kit for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient said kit comprising: 1) - optionally, instructions to apply the method or the use of claims 1 to for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient, to obtain a resulting value determining whether said patient presents SCZ, SCA, SCZall or BP disorder; and 2) a) the pair of primers depicted in Table 3 for the target IFNAR2 having the sequence SEQ ID No.6 and the sequence of its reverse primer b) optionally at least another pair, preferably 2, 3, 4, 5, 6 or the 7 pairs of primers selected from the group of pairs of primers depicted in in Table 3.

22. A kit for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient said kit comprising: 1) - optionally, instructions to apply the method or the use of claims 1 to for differential diagnosing of SCZ, SCA, SCZall or BP disorder in a patient, to obtain a resulting value determining whether said patient presents SCZ, SCA, SCZall or BP disorder; and2) a) a pair of primers allowing to obtain an amplicon including or identical to the amplicon obtainable by the pair of primers having the pair of primers depicted in Table 3 for the target AHR having the sequence SEQ ID No.1 and the sequence of its reverse primer and 5 b) optionally at least another pair, preferably 2, 3, 4, 5, 6 or the 7 pairs of primers allowing to obtain amplicon(s) including or identical to the amplicon(s) obtainable by the pairs of primers selecting from the group of pair of primers depicted in Table 3. 10

Images