Biomarkers Technical Field The present invention relates to methods of diagnosing or of monitoring psychotic disorders, in particular schizophrenic disorders and bipolar disorders, using biomarkers. The biomarkers and methods in which they are employed can be used to assist diagnosis and to assess onset and development of psychotic disorders. The invention also relates to use of biomarkers in clinical screening, assessment of prognosis, evaluation of therapy, and for drug screening and drug development Background Art The current diagnosis of psychotic conditions, such as schizophrenia and bipolar disorder, remains subjective, not only because of the complex spectrum of symptoms and their similarity to other mental disorders, but also due to the lack of empirical disease markers. There is a great clinical need for diagnostic tests and more effective drugs to treat severe mental illnesses. Psychosis Is a symptom of severe mental illness. Although it is not exclusively linked to any particular psychological or physical state, it is particularly associated with schizophrenia, bipolar disorder (manic depression) and severe clinical depression. Psychosis is characterized by disorders in basic perceptual, cognitive, affective and judgmental processes. Individuals experiencing a psychotic episode may experience hallucinations (often auditory or visual hallucinations), hold paranoid or delusional beliefs, experience personality changes and exhibit disorganised thinking (thought disorder). This is sometimes accompanied by features such as a lack of insight into the unusual or bizarre nature of their behaviour, difficulties with social interaction and impairments in carrying out the activities of daily living. Psychosis is not uncommon in cases of brain injury and may occur after drug use, particularly after drug overdose or chronic use; certain compounds may be more likely to induce psychosis and some individuals may show greater sensitivity than others. The direct effects of hallucinogenic drugs are not usually classified as psychosis, as long as they abate when the drug is metabolised from the body. Chronic psychological stress is also known to precipitate psychotic states, however the exact mechanism is uncertain. Psychosis triggered by stress in the absence of any other mental illness is known as brief reactive psychosis. Psychosis is thus a descriptive term for a complex group of behaviours and experiences. Individuals with schizophrenia can have long periods without psychosis and those with bipolar disorder, or depression, can have mood symptoms without psychosis. Hallucinations are defined as sensory perception in the absence of external stimuli. Psychotic hallucinations may occur in any of the five senses and can take on almost any form, which may include simple sensations such as lights, colours, tastes, smells) to more meaningful experiences such as seeing and interacting with fully formed animals and people, hearing voices and complex tactile sensations. Auditory hallucination, particularly the experience of hearing voices, is a common and often prominent feature of psychosis. Hallucinated voices may talk about, or to the person, and may involve severaf speakers with distinct personas. Auditory hallucinations tend to be particularly distressing when they are derogatory, commanding or preoccupying. Psychosis may involve delusional or paranoid beliefs, classified into primary and secondary types. Primary delusions are defined as arising out-of-the-blue and not being comprehensible in terms of normal mental processes, whereas secondary delusions may be understood as being Influenced by the person's background or current situation, i.e. represent a delusional interpretation of a "real" situation. 3 Thought disorder describes an underlying disturbance to conscious thought and is classified largely by its effects on the content and form of speeclh and writing. Affected persons may also show pressure of speech (speaking incessantly and quickly), derailment or flight of ideas (switching topic mid-sentence or inappropriately), thought blocking, rhyming or punning. Psychotic episodes may vary in duration between individuals. In brief reactive psychosis, the psychotic episode is commonly related directly) to a specific stressful life event, so patients spontaneously recover normal functioning, usually within two weeks. In some rare cases, individuals may refnain in a state of full blown psychosis for many years, or perhaps have attenuated psychotic symptoms (such as low intensity hallucinations) present at most times. Patients who suffer a brief psychotic episode may have many of the same symptoms as a person who is psychotic as a result of (for example) schizophrenia, and this fact has been used to support the nation that psychosis is primarily a breakdown in some specific biological system in the) brain. Schizophrenia is a major psychotic disorder affecting up to 1% of the population. It is found at similar prevalence in both sexes) and is found throughout diverse cultures and geographic areas, The World Health Organization found schizophrenia to be the world's fourth feeding cause of disability that accounts for 1.1% of the total DALYs (Disability Adjusted Life Years) and 2.8% of YLDs (years of life lived with disability). It was estimated that the economic cost of schizophrenia exceeded US$ 19 billion in 1991, more than the total cost of all cancers in the United States. Effective treatments used early in the course of schizophrenia can improve prognosis and help reduce the costs associated with this illness. The clinical syndrome of schizophrenia comprises discrete clinical features Including positive symptoms (hallucination, delusions, disorganisation of thought and bizarre behaviour); negative symptoms (loss of motivation, restricted range of emotional experience and expression and reduced hedonic capacity); and 4 cognitive impairments with extensive variation between individuals. No single symptom is unique to schizophrenia and/or is present in every case. Despite the lack of homogeneity of clinical symptoms, the current diagnosis and classification of schizophrenia is still based on the clinical symptoms presented by a patient. This is primarily because the aetiology of schizophrenia remains unknown (in fact, the aetiology of most psychiatric diseases is stil unclear) and classification based on aetiology is as yet not feasible, The clinical symptoms of schizophrenia are often similar to symptoms observed in other neuropsychiatric and neurodevelopmental disorders. Due to the complex spectrum of symptoms presented by subjects with schizophrenic disorders and their similarity to other mental disclrders, current diagnosis of schizophrenia is made on the basis of a complicated clinical examination/interview of the patient's family history, personal History, current symptoms [mental state examination) and the presence/absence of other disorders. This assessment allows a "most likely" diagnosis to tye established, leading to the initial treatment plan. To be diagnosed with schizophrenia, a patient (with few exceptions) should have psychotic, "loss-of-reallity" symptoms for at least six months (DSM IV) and show increasing difficulty in functioning normally. The JCD-10 Classification of Mental and Behavioural Disorders, published by the World Health Organization in 1992, is the manual most commonly used by European psychiatrists to diagnose mental health conditions The manual provides detailed diagnostic guidelines and defines the various forms of schizophrenia: schizophrenia, paranoid schizophrenia, hebrephrenlc schizophrenia, catatonic schizophrenia, undifferentiated schizophrenia, postschizophrenic schizophrenia, residual schizophrenia and simple schizophrenia. The Diagnostic and Statistical Manual of Mental Disorders fourth edition (DSM IV) published by the American Psychiatric Association, Washington D.C., 1994, has proven to be an authoritative reference handbook for health professionals both in the United Kingdom and in tie United States for categorising and 5 diagnosing mental health problems. This describes the diagnostic criteria, subtypes, associated features and criteria for differential diagnosis of mental health disorders, including schizophrenia, bipolar disorder and related psychotic disorders. DSM IV Diagnostic criteria for Schizophrenia A. Characteristic symptoms: Two (or more) of the following, each present for a significant portion of time during a 1-month period (or less if successfully treated): delusions, hallucinations, disorganized speech (e.g., frequent derailment or Incoherence), grossly disorganized or catatonlic behaviour, negative symptoms, i.e., affective flattening, alogia, or avolition. Only one Criterion A symptom is required if delusions are bizarre or hallucinations consist of a voice keeping up a running commentary on the person's! behaviour or thoughts, or two or more voices conversing with each other. B. Social/occupational dysfunction: For a significant portion of the time since the onset of the disturbance, one or more major areas of functioning such as work, interpersonal relations, or self-care are markedly below the level achieved prior to the onset (or when the onset is in childhood or adolescence, failure to achieve expected level of interpersonal, academic, or occupational achievement), C. Duration: Continuous signs of the disturbance persist for at toast 6 months. This 6-month period must include at least 1 month of symptcfms (or less if successfully treated) that meet Criterion A (i.e., active-phase symptoms) and may include periods of prodromal or residual symptoms. During these prodromal or residual periods, the signs of the disturbance may|be manifested by only negative symptoms or two or more symptoms listed in Criterion A present in an attenuated form (e.g., odd beliefs, unusual perceptual experiences). D. Schizoaffectlve and Mood Disorder exclusion: Schlzoaff^ctive Disorder and Mood Disorder With Psychotic Features have been ruleft out because either (1) no Major Depressive Episode, Manic Episode, or Mixe<|l Episode have occurred concurrently with the active-phase symptoms; or (2) !f fnood episodes 6 have occurred during active-phase symptoms, their total duration h£s been bn'ef relative to the duration of the active and residual periods. E. Substance/general medical condition exclusion: The disturbance is not due to the direct physiological effects of a substance (e.g., a drug of abuse, a medication) or a general medical condition, so-called "organic" brain disorders/syndromes. F, Relationship to a Pervasive Developmental Disorder: If thefle is a history of Autistic Disorder or another Pervasive Developmental Disorder, Ihe additional diagnosis of Schizophrenia is made only If prominent delusions or hallucinations are also present for at least a month (or less rf successfully treated}). Schizophrenia Subtypes 1. Paranoid Type: A type of Schizophrenia in which the following criteria are met; preoccupation with one or more delusions (especially with persecutory content) or frequent auditory hallucinations. None of the following is prominent: disorganized speech, disorganized or catatonic behaviour- or flat or inappropriate affect. 2. Catatonic Type: A type of Schizophrenia In which the clirjlcal picture is dominated by at least two of the following: motoric immobility as| evidenced by catalepsy (including waxy flexibility) or stupor excessive motor Activity (that is apparently purposeless and not influenced by external stimuli), extreme negativism (an apparently motiveless resistance to all "instructions or maintenance of a rigid posture against attempts to be movejd) or mutism, peculiarities of voluntary movement as evidenced by postufing (voluntary assumption of inappropriate or bizarre postures), stereotype^ movements, prominent mannerisms, or prominent grimacing echolalia or echobraxia, 3. Disorganized Type; A type of Schizophrenia in which the fallowing criteria are met: ad of the following are prominent: disorganized speecip, disorganized behaviour, flat or inappropriate affect. The criteria are not met fdr the Catatonic Type. 4. Undifferentiated Type: A type of Schizophrenia In which symptoms that meet Criterion A are present, but the criteria are not met foj the Paranoid, Disorganized, or Catatonic Type. 7 5. Residual Type : A type of Schizophrenia in which the following criteria are met: absence of prominent delusions, hallucinations, disorganizedl speech, and grossly disorganized or catatonic behaviour, There is continuing evidence of ths disturbance, as indicated by the presence of negative symptdms or two or more symptoms listed in Criterion A for Schizophrenia, present in Jin attenuated form (e.g., odd beliefs, unusual perceptual experiences). Schizophrenia associated features Features associated with schizophrenia Include: learning problems, hypoactivlty, psychosis, euphoric mood, depressed mood, somitjc or sexual dysfunction, hyperactivity, guilt or obsession, sexually devialit behaviour, odd/eccentric or suspicious personality, anxious or fearful pr dependent personality, dramatic or erratic or antisocial personality. Many disorders have similar or even the same symptoms as Schizophrenia: psychotic disorder due to a general medical condition, delirium or dementia; substance-induced psychotic disorder; substance-induced deliriufn; substanceinduced persisting dementia; substance-related disorders: mood disorder with psychotic features; schizoaffective disorder; depressive disorder not otherwise specified; bipolar disorder not otherwise specified; mood disordei) with catatonic features; schizophreniform disorder; brief psychotic disorder; delusional disorder; psychotic disorder not otherwise specified; pervasive .developmental disorders (e.g., autistic disorder); childhood presentations combining disorganized speech (from a communication disorder) ana disorganized behaviour (from attention-deffclt/hyperactivity disorder); schizcjtypal disorder; schizoid personality disorder and paranoid personality disorder. DSM IV Diagnostic categories for manic depression/Bipolar affective disorder (BD) Only two sub-types of bipolar illness have been defined clearty enough to be given their own DSM categories, Bipolar I and Bipolar II. 8 Bipolar I: This disorder Is characterized by manic episodes; tha 'high' of the manic-depressive cycle. Generally this manic period is followed tty a period of depression, although some bipolar I individuals may not experience a major depressive episode. Mixed states, where both manic or hypomaihio symptoms and depressive symptoms occur at the same time, also occur frequently with bipolar I patients (for example, depression with the racing thoughts of mania). Also, dysphoric mania is common, this is mania characterized by anger and irritability. Bipolar II: This disorder is characterized by major depressive episodes alternating with episodes of hypomania, a milder form of manifeu Hypomanlc episodes can be a less disruptive form of mania and may be characterized by low-level, non-psychotic symptoms of mania, such as increased energy or a more elated mood than usual. It may not affect an individual's ability to function on a day to day basis. The criteria for hypomania differ from thlose for mania only by their shorter duration (at least 4 days instead of 1 we^k) and milder severity (no marked impairment of functioning, hospftalization or psychotic features). If alternating episodes of depressive and manic symptoms lastl for two years and do not meet the criteria for a major depressive or a manic episode then the diagnosis is classified as a Cyclothymia disorder, which Is a less (severe form of bipolar affective disorder. Cyclothymic disorder is diagnosed oveh the course of two years and is characterized by frequent short periods of hjvpomania and depressive symptoms separated by periods of stability. Rapid cycling occurs when an individual's mood fluctuates frorrj depression to hypomania or mania in rapid succession with little or no period^ of stability in between. One is said to experience rapid cycling when one h^s had four or more episodes, in a given year, that meet criteria for major depressive, manic, mixed or hypomanlc episodes. Some people who rapid cycle ?an experience monthly, weekly or even daily shifts in polarity (sometimes called ultra rapid cycling). When symptoms of mania, depression, mixed mood, or hypomanla are caused directly by a medical disorder, such as thyroid disease or a strokp, the current diagnosis is Mood Disorder Due to a General Medical Condition. If a manic mood is brought about through an antidepressant. ECT :r through an individual using "street" drugs, the diagnosis is Substance-induced Mood Disorder, with Manic Features. Diagnosis of Bipolar Itl has been used to categorise manic episodes which occur as a result of taking an antidepressant medication, rather |ban occurring spontaneously. Confusfngly, it has also been used in instances where an individual experiences hypomania or cyclothymia (i.e. less severe mania} without major depression. Mania Manic Depression is comprised of two distinct and opposite sfates of mood, whereby depression alternates with mania. The DSM IV givefc a number of criteria that must be met before a disorder is classified as mania. The first one is that an individual mood must be elevated, expansive or irritable. The mood must be a different one to the individual's usual affective state difring a period of stability. There must be a marked change over a significant period of time. The person must become very elevated and have grandiose ideas. They may also become very irritated and may well appear to be 'arrogant* irl manner. The second main criterion for mania emphasizes that at least three of the following symptoms must have been present to a significant degree: inflated sense of self importance, decreased need for sleep, increased talkativeness, flight of ideas or racing thoughts, easily distracted, increased goal-directed activity. Excessive involvement in activities that can bring pleasure but may flave disastrous consequences (e.g. sexual affairs and spending excessively). The third criterion for mania in the DSM IV emphasizes that the change ifi mood must be marked enough to affect an Individual's job performance or ability to take part in 10 regular social activities or relationships with others. This third criterion is used to emphasize the difference between mania and hypomania. Depression The DSM IV states that there are a number of criteria by which major depression is clinically defined. The condition must have been [evident for at least two weeks and must have five of the following symptoms a depressed mood for most of the day, almost every day, a loss of interest pr pleasure in almost all activities, almost every day, changes in weight and Jfppetite, sleep disturbance, a decrease in physical activity, fatigue and loss of energy, feelings of worthlessness or excessive feelings of guilt, poor concentration levels, suicidal thoughts. Both the depressed mood and a loss of interest In everyday ac1|vlties must be evident as two of the five symptoms which characterize a major depression. It is difficult to distinguish between the symptoms of an individuall suffering from the depressed mood of manic depression and someone suffering from a major depression. Dysthymia is a less severe depression than unipolar depression, but it can be more persistent The prolonged process currently needed to achieve accurate diagnosis of psychotic disorders may cause delay of appropriate treatment, vthlch Is likely to have serious Implications for medium to long-term disease Outcome. The development of objective diagnostic methods, tests and tobls is urgently required to help distinguish between psychiatric diseases with! similar clinical symptoms. Objective diagnostic methods and tests for psycfiotic disorders, such as schizophrenia and/or bipolar disorder, will assist! in monitoring individuals over the course of illness (treatment response, compliance etc,} and may also be useful in determining prognosis, as well as providing tools for drug screening and drug development. Unfortunately, at present there are no standard, sensitive, specific tests for psychotic disorders, such as schizophrenia or bipolar disorders. 11 One biochemical test currently under development for schizophrenia diagnosis is the niacin skin flush test, baaed on the observation that therej is failure to respond to the niacin skin test in some schizophrenia patients, dud to abnormal arachidonfc acid metabolism. However, the specificity and sensitivfiy of this test shows an extreme inconsistency between studies, ranging from 123% to 87%, suggesting that the reliability and validity of this test still need to be| verified. International Patent Application Publication No. WO 01/632135 describes methods and compositions for screening, diagnosis, and determirjing prognosis of neuropsychiatric or neurological conditions (including BAD (bipolar affective disorder), schizophrenia and vascular dementia), for monitoring the effectiveness of treatment in these conditions and for use in drug development. Other techniques such as magnetic resonance imaging or positron emission tomography based on subtle changes of the frontal and temporal! lobes and the basal ganglia are of little value for the diagnosis, treatment, of prognosis of schizophrenic disorders in individual patients, since the absolute size of these reported differences between individuals with schizophrenic and normal comparison subjects has been generally small, with notable overlap between the two groups. The role of these neurofmaging techniques is restricted largely to the exclusion of other conditions which may be acdrampanied by schizophrenic symptoms, such as brain tumours or haemorrhages. Therefore, a need exists to Identify sensitive and specific fjiomarkers for diagnosis and for monitoring psychotic disorders, such as schizophrenic or bipolar disorders in a living subject. Additionally, there is a clear need for methods, models, tests and tools for identification and assessment of existing and new therapeutic agents for the treatment of these disorders Biomarkers present in readily accessible body fluids, such as cerebrosplnal fluid gnition (PR), expert systems and other chemoinformatic tools to interpret and classify complex NMR-generated metabolic data sets and to extract ufeeful biological information. Biofluids often exhibit very minor changes in metabolite profile! 'n response to external stimuli. Dietary, diumal and hormonal variations majf also influence biofluid compositions, and it is clearly important to differentiate} these effects if correct biochemical Inferences are to be drawn from their analysis, Biomarker information provided by NMR spectra of biofluids Is very subtly as hundreds of 13 compounds representing many pathways can often b$ measured simultaneously. ^H NMR spectra of biological samples provide a characteristic metabolic "fingerprinf or profile of the organism from which the sample was obtained for a range of biologically-important endogenous metabolites [1 - 5J, this metabolic profile is characteristically changed by a disease, disorder, toxic process, or xenobiotic (e.g. drug substance). Quantifiable differences in metabolite patterns in biological samples can give information and insight into the underlying molecular mechanisms of disease or disorder. In the evaluation off the effects of drugs, each compound or class of compound produces characteristic changes in the concentrations and patterns of endogenous metabolite^ in bioiogical samples. The metabolic changes can be characterised using automated computer programs which represent each metabolite measured in the bid.loa.ical sample as a co-ordinate in multi-dimensional space. Metabonomic technology has been used to identify biomarkers 4if inborn errors of metabolism, liver and kidney disease, cardiovascular disease, Insulin resistance and neurodegenerative disorders [3, 4, 6 - 9]. Although a wealth of disease studies have been performed on blofluids such as urirjie and plasma, relatively few metabolite profiling studies have been performed Ion CSF for the purposes of disease diagnosis and identification of key netabolHes as bfomarkers[10-15], Disclosure of the Invention In one aspect, the invention provides a method of diagnosing or monitoring a psychotic disorder In a subject comprising: (a) providing a test biological sample from said subject, (b) performing spectral analysis on said test biological sample tb provide one or more spectra, and, 14 (c) comparing said one or more spectra with one or more control spectra. Biological samples that may be tested in a method of the invention include whole blood, blood serum or plasma, urine, saliva, cerebrospinal fluid (CSF) or other bodily fluid (stool, tear fluid, synovial fluid, sputum), bijeath, e.g. as condensed breath, or an extract or purification therefrom, or dilution thereof. Biological samples also include tissue homogenates, tissue sections and biopsy specimens from a live subject, or taken post-mortem. The sajnples can be prepared, for example where appropriate diluted or concentrated and stored in the usual manner, In one embodiment, the invention provides a method of diagnosing or monitoring a psychotic disorder in a subject comprising: (a) providing a test sample of CSF from said subject, (b) performing spectral analysis on safd CSF test sample to flrovlde one or more spectra, and, (c) comparing said one or more spectra with one or more control ftpectra. Monitoring methods of the invention can be used to monitor onselt, progression, stabilisation, amelioration and/or remission of a psychotic disorder. The term "diagnosis" as used herein encompasses identificatlor and/or characterisation of a psychotic disorder, in particular a , confirmation, schizophrenic disorder, bipolar disorder, related psychotic disorder, or predisposition thereto. By predisposition it is meant that a subject does not currently present with the disorder, but is liable to be affected by the disorder in time. A psychotic disorder is a disorder in which psychosis is a reoogrlised symptom, this includes neuropsychlatric (psychotic depression and cjtier psychotic episodes) and neurodevelopmental disorders (especially Aujlstfc spectrum disorders), neurodegenerative disorders, depression, manial and in particular, schizophrenic disorders (paranoid, catatonic, disorganized, undifjferentiated and residual schizophrenia) and bipolar disorders. IS The term "biomarker" means a distinctive biological or biologically derived indicator of a process, event, or condition. Blomarkers can be used In methods of diagnosis (e.g. clinical screening), prognosis assessment; in monitoring the results of therapy, identifying patients most likely to respond tc a particular therapeutic treatment, In drug screening and development. Biotnarkers are valuable for use In identification of new drug treatments and for discovery of new targets for drug treatment. A number of spectroscopic techniques can be used to generate the spectra, including NMR spectroscopy and mass spectrometry. In preferred methods, spectral analysis Is performed by NMR spectroscopy, preferably 1H NMR spectroscopy. One or more spectra may be generated, a suite of spectra (i.e., multiple spectra) may be measured, including one for small mblecules and another for macromolecule profiles. The spectra obtained may be) subjected to spectral editing techniques. One or two-dimensional NMR spedroscopy may be performed. An advantage of using NMR spectroscopy to study complex biornflxtures is that measurements can often be made with minimal sample preparation (usually with only the addition of 5-10% DaO) and a detailed analytteal profile can be obtained on the whole biological sample. Sample volumes are small, typically 0.3 to 0.5 mi for standard probes, and as low as 3 ul for microprobes. Acquisition of simple NMR spectrji is rapid and efficient using flow-injection technology. It is usually necessary ty suppress the water NMR resonance. High resolution NMR spectroscopy (in particular 1H NMR) is particularly appropriate. The main advantages of using 1H NMR spectroscopy are the speed of the method (with spectra being obtained in 5 to 101 minutes), the requirement for minimal sample preparation, and the fact that it provides a nonselective detector for alt metabolites in the biofluid regardless of I their structural 16 type, provided only that they are present above the detection limit of the NMR experiment and that they contain non-exchangeable hydrogen atomfe. NMR studies of biological samples, e.g. body fluids, should ideally tje performed at the highest magnetic field available to obtain maximal dispersion and sensitivity and most 1H NMR studies are performed at 400 MHz or greater, e.g. 600 MHz, Usually, to assign 'H NMR spectra, comparison is made with control spectra of authentic materials and/or by standard addition of an authentic reference standard to the sample. The control spectra employed may be nprmal control spectra, generated by spectral analysis of a biological sample e,g., a CSF sample) from a normal subject, and/or psychotic disorder control spectra, generated by spectral analysis of a biological sample, (e.g., a (|;SF sample), from a subject with a psychotic disorder. Additional confirmation of assignments is usually sought from the (application of other NMR methods, including, for example, 2-dimensional (2D) NJMR methods, particularly COSY (correlation spectroscopy), TOCSY (total correlation spectroscopy), inverse-detected heteronuclear correlation metHods such as HMBC (heteronuclear multiple bond correlation), HSQC (heterohuclear single quantum coherence), and HMQC (heteronuclear multiple quantui|n coherence), 2D J-resolved (JRES) methods, spin-echo methods, relaxation editing, diffusion editing (e.g., both 1D NMR and 2D NMR such as diffusion-edited) TOCSY), and multiple quantum filtering. By comparison of spectra with normal and/or psychotic disorder cjontrol spectra, the test spectra can be classified as having a normal profile and |or a psychotic disorder profile. Comparison of spectra may be performed on entire spectra <|>r on selected regions of spectra. Comparison of spectra may involve an assessment of the variation in spectral regions responsible for deviation from the rlormal spectral 17 profile and in particular, assessment of variation in biomarkers within those regions. A limiting factor in understanding the biochemical Information froml both 1D and 2D-NMR spectra of biofluids, such as CSF, is their complexity. Although the utility of the rnetabonomic approach !s well established, its full potential has not yet been exploited. The metabolic variation ts often subtle, fend powerful analysis methods are required for detection of particular analytic, especially when the data (e. g., NMR spectra) are so complex. The most efficient way to compare and investigate these complex multiparametric data is ejmploy the 1D and 2D NMR rnetabonomic approach in combination with computer-based "pattern recognition" (PR) methods and expert systems. Metabonomics methods (which employ rnultivariate statistical analysis and pattern recognition (PR) techniques, and optionally data filtering Ijechnlques) of analysing data (e.g. NMR spectra) from a test population Uield accurate mathematical models which may subsequently be used to classify] a test sample or subject, and/or in diagnosis. Comparison of spectra may Include one or more chemometric afialyses of the spectra. The term "chemometrics" is applied to describe the i|se of pattern recognition (PR) methods and related rnultivariate statistical Approaches to chemical numerical data. Comparison may therefore comprise one or more pattern recognition analysis method(s), which can be performed b|y one or more supervised and/or unsupervised method(s). Pattern recognition (PR) methods can be used to reduce the corr|plexity of data sets, to generate scientific hypotheses and to test hypotheses, (n general, the use of pattern recognition algorithms allows the identification, ahd, with some methods, the interpretation of some non-random behaviour in a complex system which can ba obscured by noise or random variations fn ttfe parameters defining the system, Also, the number of parameters used canl be very large 18 such that visualisation of the regularities or irregularities, which foil the human brain is best in no more than three dimensions, can be difficult. Usually the number of measured descriptors is much greater than tjiree and so simple scatter plots cannot be used to visualise any similarity ar disparity between samples, Pattern recognition methods have been usefcl widely to characterise many different types of problem ranging for example over linguistics, fingerprinting, chemistry and psychology. In the context of the methods described herein, pattern recognition |is the use of muitivariate statistics, both parametric and non-parametric, to analyse spectroscopic data, and hence to classify samples and to predict the value of some dependent variable based on a range of observed measurements. There are two main approaches. One set of methods is termed "unsup|ervised" and these simply reduce data complexity in a rational way and also produce display plots which can be interpreted by the human eye. The other approach is termed "supervised" whereby a training set of samples with kn subtle expert systems may be necessary, for example, using techniques such r monitoring a subject having a psychotic disorder comprising: 22 (a) providing a test sample of CSF from said subject, (b) performing spectral analysis on said CSF test sample to prck/ide one or more spectra, (c) analysing said one or more spectra to detect the level of qne or more biomarkers present in said one or more spectra, and, (d) comparing the amount of said one or more biomarker(s) in satdione or more spectra with one or more control spectra. in particularly preferred methods, spectral analysis is performed by NMR spectroscopy, preferably 1H NMR spectroscopy. In methods of the Invention Involving spectral analysis, this may f>e performed to provide spectra from biological samples, such as CSF samples, taken on two or more occasions from a test subject Spectra from biological samples taken on two or more occasions from a test subject can be compared to identify differences between the spectra of samples taken on different occasions. Methods may include analysis of spectra from biological samples] taken on two or more occasions from a test subject to quantify the level of one or more biomarker(s) present in the biological samples, and comparing ttie level of the one or more biomarker{s) present in samples taken on two or morfe occasions. Diagnostic and monitoring methods of the invention are useful Pn mettiods of assessing prognosis of a psychotic disorder, in methods of monitoring efficacy of an administered therapeutic substance in a subject having suspected of having, or of being predisposed to, a psychotic disorder and fn methods of identifying an anti-psychotic or pro-psychotic substance. Such methods may comprise comparing the level of the one or more biomarker(s) in a biological sample, such as a CSF sample, taken from a test subject with th|e level present in one or more sample(s) taken from the test subject prior to administration of the substance, and/or one or more samples taken from the tesjt subject at an earlier stage during treatment with the substance. Additionally, Ithese methods may comprise detecting a change In the level of the on© or morelbiomarker(s) in 23 biological samples, such as CSF samples, taken from a test subject on two or more occasions, In methods of the invention in which spectral analysis is employed,{suitably one or more biomarker is selected from the group consisting of gludose, lactate, acetate (acetate species), aianine, glutamine or pH. These biomarkers of psychotic disorder, in particular schizophrenic disorder, were identified by extensive metabolic profiling analysis of CSF Samples from control and schizophrenia subjects using 1H NMR spectroBcopy ir} combination with computerised pattern recognition analysis. Significant differences in these biomarkers were found in samples obtained from first-onset, drug-fiaive patients with a diagnosis of paranoid schizophrenia when compared to age-matched normal controls. In the group with psychotic disorder, the level of glucose in CSF was found to be higher than in CSF from normal individuals; serum glucose levels were not found to be elevated in individuals \fvith psychotic disorder. The levels of lactate and acetate (acetylated species) (were found to be lower in CSF from individuals with psychotic disorder when compared to the levels in CSF from normal subjects. The pH of CSF from subjects with psychotic disorder was found on average to be 0.1 units lower jhan the pH of CSF from normal individuals. This difference in pH resulted in a) chemical shift in glutamine and aianine resonances. These differences constitute metabolic biomarkers in CSF that enable differentiation between normal individuals and those with a psychotic disorder. In an further aspect, the invention provides a method of diagnosing or monitoring a psychotic disorder, or predisposition thereto, comprising measuring the level of one or more biomarker(s) present in a ceifsbrospinal fluid sample taken from a test subject, said biomarker being selected! from the group consisting of. glucose, lactate, acetate species and pH. Such methods can be used In methods of monitoring efficacy of a therapy (e.g.) a therapeutic substance) in a subject having, suspected of having, or of being (predisposed to, a psychotic disorder. 24 Methods of diagnosing or monitoring according to the invention, rrjay comprise measuring the level of one or more of the biomarker(s) present in (fcSF samples taken on two or more occasions from a test subject. Comparisons may be made between the level of biomarker(s) in samples taken on two or more occasions. Assessment of any change in the level of biomarkeir in samples taken on two or more occasions may be performed. Modulation of the biomarker level is useful as an indicator of the state of the psychotic disorder or predisposition thereto. An increase in the level of glucose in CSF over time Is indicative of onset or progression, i.e. worsening of the disorder, whereas a decrease in the level of glucose indicates amelioration or remission of the disorder. A decrease in the level of lactate, acetylated species or pH in CSF over time is indicative of onset or progression, i.e. worsening of the disordef, whereas an increase in the level of these biomarkers indicates amelioration dr remission of the disorder. A method according to the invention may comprise comparing the level of one or more biomarker(s) in a CSF sample taken from a test subject with the level of the one or more biomarker(s) present in one or more sample(s) |aken from the test subject prior to commencement of a therapy, and/or one or njiore sample(s) taken from the test subject at an earlier stage of a therapy, JThe level of a particular biomarker is compared with the level of the same ftiomarker in a different sample, i.e. congenlc biomarfcers are compared. Such| methods may comprise detecting a change in the amount of the one or more| bjomarkers in samples taken on two or more occasions. Methods of thej invention are particularly useful in assessment of anti-psychotic therapies, in particular in drug naive subjects and In subjects experiencing their first psychotfc episode. As described herein, using methods of the invention short-term treatment with atypical anti-psychotic medication was found to result in a normalization of the disease signature in half the patients who had been commence^ on medication 25 during their first psychotic episode, whilst those who had only been) treated after several episodes did not show a normalization in CSF metabolite pfofile. A method of diagnosis of or monitoring according to the invention nhay comprise quantifying the one or more biomarker(s) in a test C8F sample tak^n from a test subject and comparing the level of the one or more biomarker(s) present in said test sample with one or more controls. The control can be selected from a normal control and/or a psychotic disorder control. The control used in a method of the invention can be one or more controls) selected frpm the group consisting of. the level of biomarker found in a normal control sample from a normal subject, a normal biomarker level; a normal biomarker rangje, the level in a sample from a subject with a schizophrenic disorder, bipolar disorder, related psychotic disorder, or a diagnosed predisposition thereto; a Schizophrenic disorder marker level, a bipolar disorder marker level, a related psychotic disorder marker level, a schizophrenic disorder marker range, a bibolar disorder marker range and a related psychotic disorder marker range. Biological samples such as CSF samples, can be taken at intervals over the remaining life, or a part thereof, of a subject. Suitably, the time elapsed between taking samples from a subject undergoing diagnosis or (nonitoring will be 3 days, 5 days, a week, two weeks, a month, 2 months, 3 months, 6 or 12 months. Samples may be taken prior to and/or during and/or following an antipsychotic therapy, such as an anti-schizophrenic or anti-bipolar disorder therapy. Measurement of the level of a biomarker can be performed b\/ any method suitable to Identify the amount of the biomarker in a CSF sampl«jj taken from a patient or a purification of or extract from the sample or a dilution thereof. In methods of the invention, quantifying may be performed by Measuring the concentration of the biomarker(s) in the sample or samples. In tjnethods of the invention, in addition to measuring the concentration of the bionjiarker in CSF, the concentration of the biomarker may be tested in a different biological sample taken from the test subject, e.g. whole blood, blood serunli, urine, saliva, or other bodily fluid (stool, tear fluid, synovial fluid, sputum), (breath, e.g. as 26 condensed breath, or an extract or purification therefrom, or dilution thereof. Biological samples also include tissue homogenates, tissue sections and biopsy specimens from a live subject, or taken post-mortem. The samples can be prepared, for example where appropriate diluted or concentrated, and stored in the usual manner. Measuring the level of a biomarker present in a sample may include determining the concentration of the biomarker present in the sample, e.g. determining the concentration of one or more metabolite biomarkfer(s) selected from glucose, acetate (acetate species) and lactate. The concentration of hydrogen ions may be measured to provide the pH value of the sjample. Such quantification may be performed directly on the sample, or indirectly on an extract therefrom, or on a dilution thereof. For example, biomarker levels can be measured by one or mq>re method(s) selected from the group consisting of: spectroscopy methods $uch as NMR (nuclear magnetic resonance), or mass spectroscopy (MS); SjELDI (-TOF), MALDI (-TOF), a 1-D gel-based analysis, a 2-D gel-based analysis, (liquid chromatography (eg, high pressure liquid chromatography (h|lPLC) or low pressure liquid chromatography (LPLC)), thin-layer chromatography, and LCMS- based techniques. Appropriate LC MS techniques include ICfAT® (Applied Blosystems, CA, USA), or ITRAQ® (Applied Biosystems, CA, UaA). Measurement of a biomarker may be performed by a direct or indirect detected, method. A biomarker may be detected directly, or indirectly, via ihteraction with a ligand or llgands, such as an enzyme, binding receptor or transporter protein, peptide, aptamer, or oligonucleotide, or any synthetic chemidal receptor or compound capable of specifically binding the biomarker. Tr|e ligand may possess a detectable label, such as a luminescent, fluorescent! or radioactive label, and/or an affinity tag. Metabolite biomarkers as described herein are suitably measured by conventional chemical or enzymatic methods (which may be direct or indirect 27 and or may not be coupled), electrochemical, fluorimetric, iLjminometric, spectrophotometric, polarimetrlc, chromatographic (e.g. HPLC) or similar techniques. For enzymatic methods consumption of a substrate in the reaction, or generation of a product of the reaction, may be detected, directly lor indirectly, as a means of measurement. Glucose can be detected and levels measured using various detection systems Including conventional chemical agents, phenylboronic acids or other synthetic receptors, or enzymatic systems, such as single enzyme systems using, for example, glucose oxldase or glucose dehydrogenase (PQQ or NAD*}; liquid chromatography, polarlmetry, refractometry, spectrophotometr|c methods, fluorlmetry, magnetic optical rotatory dispersion or near IR, an monitor the therapeutic effectiveness of existing therapies and new therapies in human subjects and in non-human animals (e.g. In animal models), Th£se monitoring methods can be incorporated Into screens for new drug substances and combinations of substances. In a further aspect the invention provides a multi-analyte panel of array capable of detecting one, two, three or four biomarker(s) selected frbm the group: glucose, acetate species, lactate, and pH. A multi-analyte panel is capable of detecting a number of different analytes. An array can be capable of detecting a single analyte in a number of samples or, as a multi-analyte array, can be capable of detecting a number of different analytes in a sample. A multi-analyte panel or multi-analyte arr^y according to 32 the invention is capable of detecting one or more metabolic |>iomarker as described herein, and can be capable of detecting a biomarker <|ir biomarkers additional to those specifically described herein. Also provided is a diagnostic or monitoring test kit suitable for performing a method according to the invention, optionally together with instructions for use of tha kit. The diagnostic or monitoring kit may comprise pne or more biosensor(s) according to the invention, a single sensor, or biosensor or combination of sensor(s) and/or biosensors may be included In the kit. A diagnostic or monitoring kit may comprise a panel or an array according to the invention. A diagnostic or monitoring kit may comprise an assay c[r combination of assays for performing a method according to the invention. Further provided is the use of one or more CSF biomarker(s) selected from glucose, lactate, acetate species, giutamine, alanine and pH to dljagnose and/or monitor a psychotic disorder. Yet further provided is tie use of a method, sensor, biosensor!, multi-analyte panel, array or kit according to the invention to identify a substance capable of modulating a psychotic disorder. A substance capable of modulating a psychotic disorder may be an anti psychotic substance useful % treatment of psychoses, or a pro-psychotic substance which may induce psychoses. Additionally provided is a method of identifying a substance capable of modulating a psychotic disorder in a subject, comprising a method of monitoring as described herein; particularly preferred identification methods comprise administering a test substance to a test subject and detecting the) level of one or more biomarker(s) selected from glucose, lactate, acetate specibs and pH in a CSF sample taken from said subject. High-throughput screening technologies based on the biomarlfers, uses and methods of the invention, e.g. configured in an array format, are suitable to monitor biomarkers for the identification of potentially useful therapeutic 33 compounds, e,Q. ligands such as natural compounds, synthetic chemical compounds (e.g. from combinatorial libraries), peptides, monoclonal or polyclonal antibodies or fragments thereof, capable of modulating the biomarker. Methods of the invention can be performed in multl-analyte pe|nel or array format, e.g. on a chip, or as a multiwell array. Methods can be adapted into platforms for single tests, or multiple identical or multiple non-ldentibal tests, and can be performed in high throughput format Methods of the Invention may comprise performing one or more additional, different tests tf> confirm or exclude diagnosis, and/or to further characterise a psychotic condilion. The identification of blomarkers for psychotic disorders, in particular schizophrenic disorders and bipolar disorders permits integration! of diagnostic procedures and therapeutic regimes. Currently there are significant delays in determining effective treatment and it has not hitherto been possible to perform rapid assessments of drug response. Traditionally, many anti-psychotic therapies have required treatment trials lasting weeks to montllis for a given therapeutic approach, Detection of biomarkers of the invention dan be used to screen subjects prior to their participation in clinical trials. Tfie biomarkers provide the means to indicate therapeutic response, failure to respond, unfavourable side-effect profile, degree of medication compliance and achievement of adequate serum drug levels. The biomarkers n|iay be used to provide warning of adverse drug response, a major problem encountered with all psychotropic medications. Biomarkers are useful in development of personalized brain therapies, as assessment of response can tye used to finetune dosage, minimise the number of prescribed medications, reduce the delay in attaining effective therapy and avoid adverse drug reactions, Thus by monitoring biomarkers in accordance with the invention, patient care can be tailored precisely to match the needs determined by the disorder and the pharmacogenomic profile of the patient; the biomarker can tflus be used to titrate the optimal dose, predict a positive therapeutic response and identify those patients at high risk of severe side effects. 34 Biomarker based tests provide a first line assessment of 'new* pjatjents, and provide objective measures for accurate and rapid diagnosis, in e| time frame and with precision, not achievable using the current subjective measures. Furthermore, diagnostic biomarker tests are useful to identify family! members or patients In the "prodromal phase", i.e. those at high risk of developing overt schizophrenia, bipolar disorder, or related psychotic disorder. This permits initiation of appropriate therapy, for example low dose anti-p^ychotics, or preventive measures, e.g. managing risk factors such as stress, illfcit drug use, or viral infections, These approaches are recognised to improve outcome and may prevent overt onset of the disorder. Biomarker monitoring methods, sensors, biosensors and kits are| also vital as patient monitoring tools, to enable the physician to determine whetner relapse is due to a genuine breakthrough or worsening of the disease, poor patient compliance or substance abuse. If pharmacological treatment ls| assessed to be inadequate, then therapy can be reinstated or increased. For genuine breakthrough disease, a change in therapy can be given if appropriate. As the biomarker is sensitive to the state of the disorder, it provides an indication of the impact of drug therapy, or of substance abuse. List of Figures Figure 1. Metabonomic analysis of CSF samples froiln drug-naive schizophrenic patients. (A) Partial 1H NMR spectrum of a CSF sample from a representative drug-nafve schizophrenia patient (grey) and a matched control (blaclt) illustrate a characteristic pH-dependent shift in the p-CHj and v-CHa Resonances of glutamine. The prominent signals at ~3.7 and 1.2ppm correspond to ethanol, a contaminant from skin disinfection prior to lumbar puncture. Theaje signals were removed from statistical analysis. 35 (B) PLS-DA scores plot showing a differentiation of drug-naive Schizophrenia patients (triangles) from demographlcaliy matched healthy voluilteer controls (squares) as determined by the 1H NMR CSF spectra. (C) PLS-DA loadings plot showing major contributing variables] towards the separation in the PLS-DA scores plots. Figure 2. Effects of "typical" and "atypical" medication on CSF metabolic profiles in first onset schizophrenia patients, (A) Spectra from a further 28 CSF samples from first onset (schizophrenia patients minimally treated (<9 days) with either typical (n«6, diamonds) or atypical (r\-22, circles) anti-psychotic medication and were compared to first onset, drug na'rve schizophrenia patients (triangles) and healjhy volunteers (squares) using PLS-DA models. The PLS-DA scores plots show that atypical anti-psychotic drug treatment resulted in a shift of approximately 50% of schizophrenia patients towards the cluster of healthy controls, (B) The same PLS-DA scores plot as (A) except that only minimally treated patients (from both drug groups) with more than one psychotic ejpisode prior to anti-psychotic treatment are shown. None of these patients shifted towards the healthy control cluster. Figure 3. Validation and prediction of schizophrenia group mempership using a PLS model. A PLS model was constructed using the OSC filtered data frorrl 37 first onset, drug naTve schizophrenia patients (empty circles) and 50 healthy volunteers (filled circles) (the 'training set'), The scores plot (A) and the leadings plot (B) indicate key resonances contributing to the separation: lattate, glucose, glutamine and citrate. This model was then used to predict "groijp membership" (i.e. disease or control) in a randomised test set of 17 first onlset, drug naive schizophrenia patients and 20 healthy volunteers which had n<|rt been used In the construction of the model. Predictions are made using a Y-pWdicted scatter plot with an a priori cut-off of 0.5 for class membership (C). 36 Figure 4. Replication of metabonomic analysis on CSF sambles from a "training sample sef comprising of 50 healthy volunteers and 3? first onset, drug naive schizophrenia patients. (A and B) PLS-DA scores and loadings plots show profiles and components discriminating between healthy volunteers (•) and drug naive schizophrenia patients (A), indicating a similar result as reported in Figure 1. Tr|ese samples Mere independently re-analyzed under an identical conditions, Noty that the key variables are highly similar to those in Figure 1. Figure 5. PLS-DA model demonstrating that gender did not influence the CSF metabolite profile in either healthy volunteers, nor in the drug naTve schizophrenia group. The symbols used are as follows: heajthy volunteer female (empty circle), healthy volunteer male (filled circle) drug naive schizophrenia female (filled triangle), drug naTve schizophrenia) male (empty triangle). Figure 8. CSF metabolite profiles of schizophrenia patients who tested positive for cannabis on urine drug screen. (A) and (B) PLS-DA scores plots showing profiles and Qiscrlmmating components of cannabis positive vs. drug naive, cannabis negative, schizophrenia patients (filled circles and triangles, respectively). (C) Localisation of cannabis positive (circles) drug naTve schizophrenia patients in the PLS-DA plot in relation to healthy volunteers (squares) Jnd drug naive schizophrenia patients who tested negative for cannabis (triangle^). Patients 153, 159 and 196 (all drug naTve schizophrenia patienls with positive urine screening for cannabinoids) show a highly altered metabolite profile (A) and appear to form a separate cluster (C). 37 Examples The invention will be further understood by reference to the examples provided below. Methods and Materials The Ethical committee of the Medical Faculty of the Unlversityl of Cologne reviewed and approved the protocol of this study and the procedures for sample collection and analysis. All study participants gave their written informed consent All clinical investigations were conducted according to (he principles expressed in the Declaration of Helsinki. CSF samples were collected from drug-naive patients diagnosed with first episode paranoid schizopHrenia or brief psychotic disorder due to duration of illness (DSM-IV 295,30 or |298.8; n=64) and from demographically matched healthy volunteers (n=7C|) (Table 1). Additionally, samples from patients fulfilling DSM-IV criteria of ((schizophrenia (DSM-IV 295.30) undergoing treatment with either typical (total n=ft: Haloperidol n=4, Perazine n»1. Fluphenazine n*1) or atypical (total n=22: Olanzapine n~9, Risperidone n-8, Quetiapine n~2, Amlsulprtde n=1, Clozapine n«1, Ziprasidone n=1) anti-psychotic medication were also included. Due to an over-representation of females in the healthy volunteer group the effect of gender on the metabolite profile was examined, but no gander-specific effect was found (Figure 5). The influence of recent and lifetime cannabis use was examined, determined by urine drug screen and clinical interview respectively (Figure 6 and Table 2). All samples were collected in a standardised fashion by the $ame team of experienced clinicians using a non-traumatic lumbar puncture procedure. Trained clinical psychiatrists performed clinical assessments. Glifcose levels in CSF and serum from healthy subjects and schizophrenic batients were measured immediately after collection using a NOVA BioProfile ajnalyser (Nova Biomedical, Waltham, USA). CSF samples were divided intd. aliquots and stored at -80'C. None of the samples underwent more than 2 freeze-thaw cycle prior to acquisition of NMR spectra. All experiments were pel-formed under 38 blind and randomized conditions. CSF samples (15Qul) were made) up to a final volume of 500pl by the addition of D2O in preparation for 1H NMR analysis. 1H NMR Spectroscopy of CSF Samples: Standard 1-D SOOMflz 1H NMR spectra were acquired for all samples using the first Increment oil the NOESY pulse sequence to effect suppression of the water resonance and lijm'rt the effect of Bo and BI inhomogeneities in the spectra (pulse sequence: relaxation delay- 90Mt-900-tm-90°-acquire FID; Bruker Analytiaohe GmbH, ^heinstetten, Germany). In this pulse sequence, a secondary radio frequency] irradiation is applied at the water resonance frequency during the relaxation de|lay of 2s and the mixing period (tm=lOOms), with ti fixed at 3us. Typically 256 transients were acquired at 300K into 32K data points, with a spectral width of GCJDOHz and an acquisition time of 1.36s per scan. Prior to Fourier transformation, the free induction decays (FID's) were multiplied by an exponential weight function corresponding to the line-broadening of 0.3Hz. Data Reduction and Pattern Recognition Procedures: To efficient!!/ evaluate the metabolic variability within and between biofluids derived frorrj patients and controls, spectra were data reduced using the software program AMIX (Analysis of Mixtures version 2.5, Bruker Rheinstetten, Germany) an.dA —, S1MCA P (version 10.5, Umetrics AB, Umea, Sweden) whe multivariate statistical analyses were conducted. Initially princip analysis (PCA) was applied to the data in order to discern exported into e a range of il components presence of inherent similarities in spectral profiles. Only one spectrum was excluded from the analysis on the basis of the Hotellings t-test which provided a 95% confidence value for a model based on the sample composition. Poor water suppression and high citrate composition were the main cause of sample exclusion. Where the classification of 1H NMR spectra was influenced by exogenous contaminants, the spectral regions containing those signals were removed from statistical analysis. In order to confirm the biomarkers differentiating between the schizophrenia patients and ma projection to latent structure discriminant analysis (PLS-DA) was ched controls, employed. 39 Orthogonal signal correction (OSC) of NMR data: The OSC met od was used to remove variation in the data matrix between samples that is with the Y-vector [16]. The resulting data set was filtered to recognition focused on the variation correlated to features of inte lot correlated allow pattern est within the sample population, this improves the predictivity and separates power of pattern recognition methods. Where appropriate, data were subjected to one-way analyst of variance (ANOVA) using the Statistical Package for Social Scientists (SPS 3/PO; SPSS, Chicago). Where the F ratio gave P<0.05, comparisons betwsen individual group means were made by Tukey's test for post-hoc comparis ons when the vanance was equal between groups. Dunnetfs T3 test was USE d for post-hoc comparisons if variances were not equal. The significance levkis was set at p=0.05. Plots of PLS-DA scores based on 1H NMR spectra of CSF sambles showed a clear differentiation between healthy volunteers and drug-na'fv first onset, paranoid schizophrenia (Figure 1). The load 3 patients with g coefficients indicated that glucose, acetate, alanine and glutamine resonances were predominantly responsible for the separation between classes. NMR spectroscopy showed significantly elevated glucose co CSF samples from first-onset, drug-naive, paranoid schizophre nia patients as compared to the demographically matched control group, increase in concentration of 6,5% ± 0.94% (p ~ 0.04, One-way / measurements of CSF glucose levels (performed immediate collection) confirmed that glucose levels in drug-naive schizophi the first cohort were significantly higher than in healthy vc increase, p=0.005; Table 1), esults from 1H centrations in Ith a relative ^OVA). Direct after sample nia patients in unteers (6.5% 40 Table 1 Demographic details, CSF and Healthy Volunteer (HV) (n=70) Drug Naive Paranoid Schizophrenia (PS, 1" cohort) (n=37) serum glucose levels of SL Drug Naive Paranoid Schizophrenia (PS 2nd l_ '* cohort) Schfzophreni a treated with "typical" anflpsychotlc (ST) (n=6) bjects Schizophrenia treated with "atypical" antlpyschotfc (SAT) (n=22) Age(yrs)1 27.4 ± 5.9 28.1 ± 9,4 25.0 ± 5,6 31 £±5.5 29.2 ± 10.1 Sex" male female [Glucose](mg/dl) CSF Serum Duration of treatment (days) 39 31 5B.5±4.6* 87.2 ±15.0** N/A 27 10 62.3 ± 5.5 93,1 i 14,4 N/A 12 5 65.3 ±6.4 91 .5 ±9.9 N/A 5 1 65,0 * 5.9 87.3 ±19.2 9.6 ± 8.3 17 5 64.9 ±6.4 103.5 ±24.7 9,2 ± 6.2 # There is no significant difference in age between the contrpl and disease groups (Oneway-ANOVA). * Female gender is over-represented in the HV group, but sex appears to have no effect on CSF metabolite profiles (see Figure 5). Glucose levels in CSF from healthy volunteers (HV) are ower than the glucose levels In CSF from drug-naive paranoid schizophrenia" patients (PS), paranoid schizophrenia patients treated with typical (ST) and atypical (SAT) anti-psychotic medication (HV vs. PS (two cohorts included), q<0.001; HV vs. SAT, p<0.001; HV vs. ST, p*0.02, One-way ANOVA wrth Tukeyi test), **Serum glucose levels are significantly increased only in schizophrenia patients treated with atypical antl-psychotlcs (HV vs. SAT, plo.05, One-way ANOVA with Dunnett's T3 test). There is no significant difference in serum glucose level between other groups. All data are shown in mean ± s. d. 41 Interestingly, serum glucose levels obtained from the same schizophrenia and healthy subjects showed no difference (p=0.24), suggesting specific elevation in glucose levels. In contrast, acetate concentrations were reduced (11.5%, p - 0.006; and 17.3%, p respectively) in drug-naive schizophrenia patients (the first cohort matched controls. Spectral changes corresponding to glutamin resulted from a pH dependent change in the chemical shift of thes The pH of CSF samples from untreated schizophrenia patients wi brain/CSFand lactate '0.05 (t test), compared to 2 and alanine 3 resonances, s found to be on average 0.1 pH units lower than in the matched control samples (p<0.05, t test) which corresponded to a mean chemical shift change of O.OiS ppm for the B-CH2 resonance of glutamine and 0.016 ppm shift change for trie alanine CH3 signal. Short term treatment for an average of nine days (see atypical anti-psychotic medication resulted in a normalisatior metabolite profile in approximately 50% of the schizophrenia p Table 1) with of the CSF itients (Figure 2A), whereas treatment with typical anti-psychotic medication did not show such an effect (Figure 2A), although as the number of patients treated with typical antl-psychotics is low (n-6), no clear conclusions can be drawn from this observation. Interestingly, it was observed that patients who suffered several psychotic episodes before drug treatment was initiated (either with typical or atypical anti-psychotics) did not show a normalisation of theiJ CSF disease profile over the duration of the study. Six out of a total of seve|n patients with more than one episode before drug treatment clustered closely naive schizophrenia group and, indeed, none of them clustered \ nth the healthy control group (Figure 2B). Moreover, all schizophrenia patients \ /ho exhibited a with the drugnormalisation of the CSF metabolite profile (either with typical psychotics) had commenced medication dun'ng their presentation. In statistical terms (recognising that numbers study implies that if treatment is initiated during a first episode, i ir atypical antiirst psychotic re small), this 7% of patients recover (assessed in terms of normalisation of CSF metabolite p 'ofiles), whilst if medication was given after a second psychotic episode, no normalisation (0/7) was observed within the time frame of this study. I 42 Due to the prevalent cannabla use amongst schizophrenia pati snts and the known influence of cannabis on glucoregulation, the influence of this potential confounding factor was examined in the disease and control gro the control patients had tested positive on urine drug screen and CSF metabolites was observed between healthy volunteers ps. None of 10 change in t/ho reported moderate (>5 times/ lifetime) or low/no (<2 times/lifetime) cannasis use (data not shown). In the drug na'fve, paranoid schizophrenia group, 7 p dents (out of a total of 37) tested positive for cannabis on urine drug scree i. Cannabis positive patients had significantly lower serum glucose levels ( p=0.05, t test), but no effect on CSF glucose levels was obsen, ed (p=0.20, t test; see Figure 6 and Table 2). Three patients who tested positiv were found to have highly altered CSF metabolite profiles « separate cluster in the PLS-DA plot (away from both healthy % decrease; for cannabis nd formed a controls and schizophrenia patients) whilst the remaining four cannabis po itlve patients clustered with the drug negative group (see Figure 6). Table 2. Effect of cannabis use on serum and CSF glucose levels in paranoid schizophrenia patients, i Paranoid schizophrenia Paranoid patients with cannabis patients "positive" In urine "negative" (n=7) (n-30) schizophrenia fth cannabis i urine CSF glucose concentration Serum glucose concentration 60.3 ± 4.3 86.3 ± 9.0 62.9 ± 5.7 95.1 ± 15.1 Data are shown as mean ± S. D, Data are shown as mean * p=0.05, t test. Validation of key metabolic alterations in an independent test f validate the findings, samples from the first cohort (70 control an drug naive schizophrenia CSF samples), were re-analyzed ampte set. To i 37 first onset, alongside a second 43 cohort of 17 additional first onset, drug naTve schizophrenia patierjts. A model was built based on a training set of 50 randomly selected control Bamples and 37 first onset, drug naTve schizophrenia samples from the first (fohort. Both PCA and PLS-DA showed similar results as shown in Figure 1 (Fijjure 4). This mode! was then used to predict class membership in a test set comprising of 20 control CSF samples (from the first cohort) and 17 first onsel schizophrenia patients (from the 2nd cohort, Table 2). Orth correction (OSC) was applied to enhance the metabolic different!! , drug naTve gonal signal tion between classes within the model [4]. After OSC, the separation of control and first onset, drug naTve schizophrenia groups in the PIS scores plots (F gure 3A) was characterized by similar spectral regions to those previously identified as contributing to the separation of the classes, i.e. glucose, lactate, shifts in glutamine resonances and citrate (Figure 3B). The PLS model calculated from OSC-filtered NMR data was then used to predict class membership in the test sample set. The Y-predicted scatter piot assigned samples to either to the control or schizophrenia group using an a priori cut-off of 0.5, arid showed the ability of 1H~NMR metabonomics analysis to predict class membership of unknown samples with a sensitivity of 82% and a specificity of 85% (Figure 3C). Analysis of the 1H NMR spectra of CSF samples showe distribution of samples from healthy volunteers away from dru with first onset schizophrenia (Figure 1B and 1C). The rnetabol' was found to be characteristically altered in schizophrenia p majority of key metabolites contributing to the separation were independent test set (Figure 3). There was some overlap of classes in the PLS-DA scores plot derived from the NMR spe and 1C). Whilst the drug nafve, paranoid schizophrenia group tightly together, a small number of samples did not show a cle the PLS-DA analysis. This may Indicate the existence of sen groups; also clinical parameters, such as disease progression, drug-response may relate to distinct metabolic signatures. Althc size of this study was too small to enable strong conclusion subgroups to be drawn, it was of interest that all 4 patients wh a differential naTve patients profile of CSF tients and the eplicated in an two sample (Figure 1B clustered very separation in zophrenia sub" severity and/or gh the sample about patient were found to 44 cluster with the control group (Figure 1B), had an exceptionally gjood outcome or recovered fully from a first episode of psychosis. Abnormal glucose levels in serum have been linked with antitreatment [17,18], yet the observations made in this study of CSF glucose concentrations In schizophrenia patients giucoregulatory alterations are intrinsic to the schizophrenia syn brain-specific, because samples collected from drug-naive, first showed significantly increased CSF glucose levels and glucose elevation was not observed in sera from the same schizophrenia subjects. Elevated CSF glucose has not previously been reported for schizophrenia, how aver abnormal fasting glucose tolerance has been observed in serum from first [19]. The prevalence of diabetes type (I is substantially schizophrenia patients (15.8% as compared to 2-3% in the gen [20], Studies have also found increased plasma levels of norepinephrine in schizophrenia patients [21-23] although in glucose and the high prevalence of type II diabetes In schizop have mainly been attributed to anti-psychotic drug treatment [17 this study, serum glucose levels were found to be increased in \ atients treated with atypical anti-psychotic medication (Table 1). It Is pos; treatment precipitates the onset of diabetes in schizophrenia context of a co-predisposition and that both schizophrenia and c share common disease mechanisms, The significantly lower CJ sychotic drug i elevation of imply that rome and are inset patients onset patients increased in ral population) glucose and eased serum renic patients 23]. Indeed, in ble that drug atients in the abetes type II pH observed aligns with observations in post-mortem brain and may b* attributed to alterations in energy metabolism at large [24]. Numerous other mortem brain have also found mitochondria! changes in sen [25.26]). The lowered pH observed in CSF in this study may alterations in cellular respiration. Surprisingly, however, whilst tudies on post- :ophrenia (e.g. :hus be due to an increase in lactate in post-mortem brain tissue has been found, in this stu iy a significant decrease in CSF lactate levels was detected in first onset schizophrenia patients. At this stage it is not possible to determine which metat olite alterations are contributing to the lowered pH in CSF. A possible explanatl< >n could be that the "schizophrenia brain" preferentially utilizes lactate over glu ;ose as energy 45 substrate. Brain lactate is believed to be predominantly produced jay astrocytes [27] and is used as energy substrate in brain, in particular by nburons under certain conditions [27]. In fact, significant monocarboxylate utilization by the brain was also reported in different pathological states such as diabetes and prolonged starvation [28,29]. Acetate was also found to be significantly reduced in the CSF of first-onset, drug na'fve schizophrenia patients. The majority of acetate ify the brain is utilised in fatty acid and lipid synthesis [30], thus the decreased acetate concentration may suggest a compromised synthesis of myefilvrelated fatty acids and lipids in the schizophrenia brain. Acetate in the bra derived from N-acetylaspartate (NAA), which is hydrolyzed into L acetate by the enzyme aspartoacylase (ASPA) [31]. NAA is in is primarily -aspartate and synthesized in neuronal mitochondria and transferred to oligodendrocytes, where ASPA liberates the acetate moiety to be used for myelin lipid synthesis 32]. An in vivo reduction in NAA levels in schizophrenia is a well-established observation [33]. More interestingly, we found ASPA transcripts down-regulated in post-mortem brain using micro-array and quantitative PCR (Q-PCR) analysis ir| schizophrenia post-mortem brain (-178; p=0.09 by microarray; - 1.61; p=0j04 by Q-PCR; n-15 schizophrenia prefronta! cortex and matched controls; unpublished). Together with our findings of a significant decrease of acetate in pSF, this (ends further support not only for altered NAA metabolism, but also for oligodendrocyte dysfunction, which we and others previously reported [34,35]. Perturbations in CSF acetate concentrations have afso beeln observed In patients with CJD, although in contrast to the current stydy, CJD was associated with an increase in acetate concentrations [36]. Disturbed glucose metabolism has also been associated \|/ith mood and psychotic disorders [37], although to our knowledge none cjf these studies measured CSF glucose levels. However, the increased concentrations of glucose together with other metabolic perturbations, such as lower levels of acetate and factate, and a pH-dependent shift in glutamine resonances, may represent a more specific disease diagnostic for schizophrenia. 46 The effects of two drug treatment regimen, the use of typical anc| atypical antipsychotic medication, were evaluated using the same analytical methods. Normalization of the metabolite profiles was observed in patients (na28) who had been treated with atypical anti-psychotic medication for an average of 9 days. Figure 2 Illustrates a shift of approximately 50% of patients on atypical anti-psychotics towards the cluster of healthy controls within the| PLS-DA plot These results are indicative that atypical medication results in a normalization of the metabonomic disease signature. It Is a well-established (fact that only between 50-70% (according to different sources) of schizophrenia patients respond to anti-psychotic intervention. However, clinical response is generally only observed after weeks or months of treatment, it is believed that normalization of the metabonomic signature detected in this study is liable to be predictive of clinical drug response. One of the most striking findings of this study is the effect of number of psychotic episodes prior to commencing anti-psychotic treatment on CSF metabolite profile in paranoid schizophrenia patients. 57% of patients who were commenced on anti-psychotic medication during their first psychotic episode were found to cluster with the healthy control cluster whereasl six out of the seven patients who had several psychotic episodes prior to treatment clustered with the drug-naTve, paranoid schizophrenia group (Figure 2B). These results suggest that the initiation of anti-psychotic treatment during a| first psychotic episode may influence treatment response or indeed outcome, fhis view is in agreement with The Personal Assessment and Crisis Evaluation (PACE) clinic study [38J, the Prevention through Risk Identification, Management and Education (PRIME) study [39] and other ongoing studies that purport that early Identification of patients at risk of developing schizophrenia vjith subsequent Intervention may reduce morbidity and adverse outcome. Metabonomic approaches to profiling CSF employed in this study provide a nfew approach to achieving both early diagnosis and monitoring therapeutic intervention for schizophrenia. 47 As many schizophrenia patients are recreational cannabis u$ers and as cannabls has a known effect on glucoregulation, this potential confounding factor was examined. Recent cannabis use was associated with] a significant reduction in serum glucose, but no influence on the CSF metabolite profile was observed. The application of metabolite profiling tools as described herslrj provides an efficient means for early diagnosis of psychotic disorders sucn as paranoid schizophrenia and provides a practical method for monitoring therapeutic intervention by providing metrics for the normalization of biofldld spectra by muitivariate comparison with the relevant control profiles. References: 1. Nicholson JK, Lindon JC, Holmes E (1999) 'Metabonomics': Understanding the metabolic responses of living systems to pathophysiologiial stimuli via muitivariate statistical analysis of biological NMR spectnpscopic data. Xenobiotica 29:1181-1189. 2. Tsang TM, Griffin JL, Haselden J, Fish C, Holmes E (2CJ05) Metabolic characterization of distinct neuroanatomical regions in rats b^ magic angle spinning (1)H nuclear magnetic resonance spectroscopy. Magn EJteson Med 53: 1018-1024. 3. Nicholson JKr Connelly J, Linden JC, Holmes E (2002) Mejtabonomics: a platform for studying drug toxicity and gene function. Nat Rev prug Discov 1: 153-161. 4. Brindle JT, Anttl Ht Holmes E, Tranter G, Nicholson JK, et a|. (2002) Rapid and noninvasive diagnosis of the presence and severity of [:oronary heart disease using 1H-NMR-based metabonomics. Nat Med 8:1439- 444. 5. Nicholson JK, Holmes E, Lindon JC, Wilson ID (2004) Thcl challenges of modeling mammalian biocomplexity. Nat Biotechnol 22:1268-121/4. 6. Cheng LL, Newell K, Mallory AE, Hyman BT, Gonzalpz RG (2002) Quantification of neurons in Alzheimer and control brains wil|h ex vivo high resolution magic angle spinning proton magnetic resonance spjectroscopy and stereology. Magn Reson Imaging 20: 527-533. 48 7. Cheng LL, Ma MJ, Becerra L, Ptak T, Tracey I, et al. (1997|> Quantitative neuropathology by high resolution magic angle spinning pro|ton magnetic resonance spectroscopy. Proc Nat! Acad Sci U S A 94:6408-6413|. 8. Beckwtth-Hall BM, Nicholson JK, Nicholls AW, Foxall PJ, Linpon JC, et al. (1998) Nuclear magnetic resonance spectroscopic and principal components analysis investigations into biochemical effects of three model hepatotoxins. Chem Res Toxicol 11: 260-272. 9. Holmes E, Foxall PJ, Spraul M, Farrant RD, Nicholson JK, etlal, (1997) 750 MHz 1H NMR spectroscopy characterisation of the complex metabolic pattern of urine from patients with inborn errors of metabolism: 2-iJiydroxyglutaric aciduria and maple syrup urine disease. J Pharm Biomed Anal 15|: 1647-1659. 10. Garseth M, Sonnewald U, White LR, Rod M, Nygaard O et al. (2002) Metabolic changes in the cerebrospinal fluid of patients witiji lumbar disc herniation or spinal stenosis. J Neurosci Res 69: 692-695. 11. Braun KP, Gooskens RH, Vandertop WP, Tulteken CA, vah der Grond J (2003) 1H magnetic resonance spectroscopy in human hydrocetfhalus. J Magn Reson Imaging 17:291-299, 12. Koschorek F, Offermann W, Stelten J, Braunsdorf WE, Stellar U, et al. (1993) High-resolution 1H NMR spectroscopy of cerebrospins" fluid in spinal diseases. Neurosurg Rev 16; 307-315. 13. Hashimoto K, Engberg G, Shimizu E, Nordin C, Lindstrom ., et al. (2005) Elevated glutamlne / glutamate ratio in cerebrospinal fluid of firjst episode and drug naive schizophrenic patients. BMC Psychiatry 5:1-6. 14. White LR, Garseth M, Aasly J, Sonnewald U (2004) CerebroioJnal fluid from patients with dementia contains increased amounts of an unknown factor. J Neurosci Res 78: 297-301, 15. Do KQ, Trabesinger AH, Kirsten-Kruger M, LauerCJ, DydaklU, et al. (2000) Schizophrenia: glutathione deficit in cerebrospinal fluid and preirontal cortex in vivo, Eur J Neurosci 12:3721-3728. 16. Wold S, Antti H, Lindgren F, Ohman J (1998) Orthogonal sfcgnal correction of near-infrared spectra. Chemometrics Intelligent Lab Systems f4: 175-185. 17. Henderson DC, Cagliero E, Copeland PM, Borba CP, Ev(ns|E, et al. (2005) Glucose metabolism in patients with schizophrenia treated wijh atypical anti49 psychotic agents: a frequently sampled intravenous glucose tolerance test and minimal model analysis. Arch Gen Psychiatry 62:19-28. 18. Newcomer JW (2004) Abnormalities of glucose metabolism associated with atypical anti-psychotic drugs. J Clin Psychiatry 65 Suppl 18:36-46 19. Ryan MC, Collins P, Thakore JH (2003) Impaired fasting glucose tolerance in first-episode, drug-naive patients with schizophrenia. Am J Psychiatry 160: 284-289, 20. Henderson DC, Ettinger ER (2002) Schizophrenia and dtatjetes, Int Rev Neurobiol 51:481-501. 21. Arranz B, Rosel P, Ramirez N, Duenas R, Fernandez P, et al, (2004) Insulin resistance and increased leptin concentrations in noncompliant schizophrenia patients but not in anti-psychotic-naive first-episode schizophrenia patients. J Clin Psychiatry 65:1335-1342. 22. Dinan T, Peveler R, Holt R (2004) Understanding schizophrenia and diabetes. Hosp Mad 65:485-488. 23. Elman I, Rott D, Green Al, Langleben DD, Lukas SE, et al. (2t)04) Effects of pharmacological doses of 2-deoxygfucose on plasma catecr olamines and glucose levels in patients with schizophrenia. Psychopharmacoligy (Bert) 176: 369-375. 24. Prabakaran S, Swatton J, Ryan M, Huffaker H, Huang TJ, e|t al. (2004) An integrative functional genomics approach reveals impaired brain energy metabolism In Schizophrenia. Mol Psychiatry: (in press). 25. Iwamoto K, Bundo M, Kato T (2005) Altered expression o\ mitochondriarelated genes in postmortem brains of patients with bipolar disorder or schizophrenia, as revealed by large-scale DNA microarray anajysis. Hum Mol Genet 14: 241-253. 26. Karry R, Klein E, Ben Shachar D (2004) Mitochondrial convex I subunits expression is altered in schizophrenia: a postmortem study. Biol) Psychiatry 55: 676-684. 27. Pierre K, Pellerin L (2005) Monocarboxylate transporters in the central nervous system: distribution, regulation and function. J Neurochejm 94:1-14, 28. Hawkins RA, Mans AM, Davis DW (1986) Regional ketone [body utilization by rat brain In starvation and diabetes. Am J Physiol 250: E169-1J78. 50 29. Fernandas J, Berger R, Smlt GP (1982) Lactate as energy scjurce for brain in glucose-6-phosphatase deficient child. Lancet 1; 113. 30. Kammula RG, Fong BC (1973) Metabolism of glucose and dcetate by the ovine brain in vivo. Am J Physiol 225:110-113. 31. Madhavarao CN, Arun P, Moffett JR, Szucs Sf Surendran SL et al. (2005) Defective N-acetylaspartate catabolism reduces brain acetate levels and myelin liptd synthesis in Canavan's disease. Proc Natl Acad Sci U S A 10(2: 5221-5226, 32. Chakraborty G, Mekala P, Yahya D, Wu G, Ledeer RW (2001) tntraneuronal N-acetylaspartate supplies acetyl groups for myelin lipid synthesis: evidence for myelin-associated aspartaacylase. J Njeurochem 78: 736-745, 33. Steen RG, Hamer RM, Lleberman JA (2005) Measurement of brain metabolites by 1H magnetic resonance spectroscopy in patients with schizophrenia: a systematic review and meta-analysis. Neuropsychopharmacology 30:1949-1962, 34. Prabakaran S, Swatton JE, Ryan MM, Huffaker SJ, Huang Jff, et al. (2004) Mitochondrial dysfunction in schizophrenia: evidence for compromised brain metabolism and oxldative stress. Mol Psychiatry 9:634-697,643 35. Hakak Y, Walker JR, Li C, Wong WH, Davis KL, et al. (2001 [ Genome-wide expression analysis reveals dysregulatbn of myelination-relfated genes in chronic schizophrenia. Proc Natl Acad Sci U S A 98:4746-4751 36. Maillet S, Vion-Dury J, Confort-Gouny S, Nicoli F, Lutz N\M et ai. (1998) experimental protocol for clinical analysis of cerebrospinall fluid by high resolution proton magnetic resonance spectroscopy. Brain ftes Brain Res ProtocS: 123-134. 37. Regenold WT, Phatak P, Kling MA, Mauser P (2004) Post-ir|oitem evidence from human brain tissue of disturbed glucose metabolism) in mood and psychotic disorders. Mol Psychiatry 9:731-733. 38. McGorry PD, Yung AR, Phillips LJ, Yuen HP, Francey J3, et al. (2002) Randomized controlled trial of interventions designed to redfuce the risk of progression to first-episode psychosis in a clinical sample witfi sub threshold symptoms. Arch Gen Psychiatry 59:921-928. 51 39. McGlashan TH. Abstract presented at the Twelfth Biennial Winder Workshop on Schizophrenia. In: Davos, editor; 2004. Switzerland. 40. Geladi, P., and 6. R. Kowalski (1986), "Partial Least Squares Regression; A Tutorial," Analytics Chimica Acta, 185,1-17, CLAIMS: 1. A method of diagnosing or monitoring a psychotic disorder in a subject comprising: (a) providing a test biological sample of CSF from said subject, (b) performing spectra) analysis on said CSF test sample to provide one or more spectra, and, (c) comparing said one or more spectra with one or more control Spectra. 2. A method according to claim 1, wherein spectral analysis is performed by NMR spectroscopy. 3. A method according to claim 1 or claim 2, wherein spectral analysis is performed by 1H NMR spectroscopy. 4. A method according to any preceding claim, wherein the one or more control spectra comprise normal control spectra. 5. A method according to any preceding claim, wherein the one or more control spectra comprise psychotic disorder control spectra. 6. A method according to any preceding claim, wherein said comparing comprises classifying spectra performed on a test CSF sample as having a normal or psychotic disorder profile. 7. A method according to any preceding claim, wherein said comparing comprises assessing variation in one or more biomarkers present in said spectra, 8. A method according to any preceding claim, wherein (said comparing comprises one or more chemometric analyses. 9. A method according to any preceding claim, wherein said comparing comprises a pattern recognition analysis. 10. A method according to any preceding claim, wherein pattern recognition analysis is performed by one or more supervised and/or unsupervised method(s). 11. A method according to claim 10, wherein the one or morel unsupervised method(s) is/are selected from: a principle components analysis (PCA), non- linear mapping (NLM) and a clustering method. 12. A method according to claim 10 or claim 11, wherein the one or more supervised method(s) is/are selected from: a soft independent modelling of class analogy, a partial least squares (PLS) method, a k-nearest neighbour analysis and a neural network. 13. A method of diagnosing or monitoring a subject having a psychotic disorder comprising: (a) providing a test sample of CSF from said subject, (b) performing spectral analysis on said CSF test sample to provide one or more spectra, (c) analysing said one or more spectra to detect the level of one or more biomarkers present in said one or more spectra, and, (d) comparing the amount of said one or more biomarker(s) in said one or more spectra with one or more control spectra, 14. A method according to any preceding claim, comprising performing spectral analyses to provide spectra from CSF samples taken pn two or more occasions from a test subject. 15. A method according to claim 14, comprising comparing spectra from CSF samples taken on two or more occasions from a test subject. 16. A method according to claim 14, comprising analysing spectra from CSF samples taken on two or more occasions from a test subject to quantity one or more biomarker(s) present in the CSF samples, and comparing the level of the one or more biomarker(s) present In CSF samples taken on two or more occasions. 17. A method of assessing prognosis of a psychotic disorder comprising 9 method according to any preceding claim. 18. A method of monitoring efficacy of a therapeutic substance in a subject having, suspected of having, or of being predisposed to, a psychotic disorder, comprising a method according to any preceding claim, 19. A method of identifying an anti-psychotic substance, comprising a method according to any preceding claim. 20. A method of identifying a pro-psychotic substance, comprising a method according to any preceding claim. 21. A method according to any one of claims 18 to 20, comprising comparing the level of the one or more biomarker(s) in a CSF sample taken from a test subject with the level present In one or more sample(s) taken from the test subject prior to administration of the substance, and/or one or more samples taken from the test subject at an earlier stage during treatment with the substance. 22. A method according to any one of claims 14 to 21, further comprising detecting a change In the level of the one or more biomarker(s) n CSF samples taken from a test subject on two or more occasions. 23. A method according to any preceding claim, wherein the biomarker is selected from glucose, lactate, acetate species, alkaline, glutamine or pH. 24. A method of diagnosing or monitoring a psychotic disorder, or predisposition thereto, comprising measuring the level of one or more biomarker(s) present in a cerebrospinal fluid sample taken from 3 test subject, said biomarker(s) being selected from the group consisting of: glucose, lactate, acetate species and pH, 25. A method of monitoring efficacy of a therapy in a subject having, suspected of having, or of being predisposed to, a psychotic disorder, comprising a method according to claim 24. 26. A method according to claim 24 or claim 25, comprising measuring the level of one or more of the biomarkar(s) present in CSF samples taken on two or more occasions from a test subject. 27. A method according to claim 26, comprising comparing the level of the one or more biomarker(s) present in CSF samples taken on two or more occasions from a test subject. 28. A method according to any one of claims 24 to 26, comprising comparing the level of one or more biomarker{s) in a CSF sample taken from a teat subject with the level of the one or more biomarker(s) present in one or more sample(s) taken from the test subject prior to commencement of a therapy and/or one or more sample(s) taken from the test subject at an earlier stage of a therapy. 29. A method according to any one of claims 24 to 28, wherein the therapy is an anti-psychotic disorder therapy. 30. A method according to any one of claims 24 to 29, comprising detecting a change in the amount of the one or more biomarker(s) in samples taken on two or more occasions. 31. A method according to any one of claims 24 to 30, comprising comparing the amount of the one or more biomarker(s) present in a CSF sample with the level of the one or more biomarker(s) in one or more control(s). 32. A method according to claim 31, wherein the control(s) are a normal control and/or a psychotic disorder control. 33. A method according to any preceding claim, wherein CSF samples are taken at intervals over the remaining life, or a part thereof, of a subject. 34. A method according to any preceding claim, comprising quantifying one or more biomarker(s) in a further biological sample taken from the test subject. 35. A method according to claim 34, wherein the further biological sample Is selected from the group consisting of: whole blood, blood serum, urine, saliva, or other bodily fluid, or breath, condensed breath, or an extract or purification there from, or dilution thereof. 36. A method according to any one of claims 24 to 35, wherein the level of one or more biomarker is detected by analysis of NMR spectra. 37. A method according to any preceding claim wherein the level of a biomarker is detected by one or more method selected from the group consisting of: NMR, SELDI (TOF) and/or MALDI (-TOF), a 1-D gel-based analysis, a 2-D gel-based analysis, mass spectrometry (MS) and LC-MS-based technique. 38. A method according to any preceding claim, wherein the level of one or more biomarkers is detected by one or more method selected from: direct or Indirect, coupled or uncoupled enzymatic methods, electrochemical, spectrophotometric, fluorimetric, luminometric, spectrometric, polarimetric and chromatographic techniques. 39. A method according to any preceding claim wherein the level of a biomarker is detected using a sensor or biosensor comprising one or more enzyme(s), binding, receptor or transporter protein(s), synthetic eceptor(s) or other selective binding molecule(s) for direct or indirect detection of the biomarker(s), said detection being coupled to an electrical, optical, acoustic, magnetic or thermal transducer. 40. A method according to any preceding claim wherein the psychotic disorder is a schizophrenic disorder. 41. A method according to claim 40, wherein the schizophrenic disorder is selected from the group consisting of: paranoid, catatonic, disorganized, undifferentiated and residual schizophrenia. 42. A method according to any preceding claim wherein the psychotic disorder is a bipolar disorder, 43. A psychotic disorder sensor or biosensor capable of quantifying one, two, three or four biomarker(s) selected from the group: glucose, lactate, acetate and PH. 44. A psychotic disorder sensor or biosensor according to claim 43, wherein the level of one, two, three or four biomarker(s) is detected by one or more method selected from: direct, indirect or coupled enzymatic, spectrophotometric, fluorimetric, luminometric, spectrometric, popterimetric and chromatographic techniques. 45. A psychotic disorder sensor or biosensor according claim 43 or claim 44, wherein the level of a biomarker is detected using sensor or biosensor comprising one or more enzyme(s), binding, receptor or transporter protein (s), synthetic receptor(s) or other selective binding molecule(s) for direct or indirect detection of the biomarker(s), said detection being coupled to an electrical, optical, acoustic, magnetic or thermal transducer, 46. An array or multi-analyte panel capable of detecting one two or more biomarker(s) selected from the group: glucose, acetate, lactate, and pH. 47. A diagnostic or monitoring kit suitable for performing a method according to any one of claims 1 to 42, optionally together with instructions for use of the kit 48. A kit according to claim 47, comprising one or more sensor(s) and/or biosensor(s) according to claims 43 to 45, optionally together with instructions for use of the kit. 49. A kit according to claim 47, comprising one or more analys(s) or multi- analyte panel(s) according to claim 46, optionally together with instructions for use of the kit. 50. A kit according to claim 47 comprising one or more assay(s), capable of detecting one, two or more biomarker(s) selected from the group: glucose, acetate, lactate, glutamine, alanine and pH. 51. The use of one or more CSF biomarker(s) selected from glucose, lactate, acetate, glutamine, alanine and pH to diagnose and/or monitor a psychotic disorder. 52. The use of a method, sensor, biosensor, multi-analyte panel, array or kit according to a preceding claim to identify a substance capable (of modulating a psychotic disorder. 53. A method of identifying a substance capable of modulating a psychotic disorder in a subject, comprising administering a test substance to a test subject and detecting the level of one or more biomarker(s) selected from glucose, lactate, acetate species and pH in a CSF sample taken from said subject