A method of diagnosing and / or differentiating a neurological disorder

Abstract

The present invention provides a method for diagnosing a patient having a neurological (including neuromuscular) disorder, which method extends to the pre-symptomatic diagnosis of said patients. The invention further relates to a method for distinguishing between neurological disorders, said method extending to include the ability to discriminate one subtype from other subtypes of the same disorder. Examples of the latter include the ability to discriminate: spinal muscular dystrophy type II (astasia) from spinal muscular atrophy type II or type IV (mobile but may subsequently lose ambulation); "fast- progression" amyotrophic lateral sclerosis from "slow-progression" amyotrophic lateral sclerosis; and "severe disease stage" amyotrophic lateral sclerosis from "mild disease stage" amyotrophic lateral sclerosis. In this way, the present invention allows one to monitor disease state progression of a neurological disorder in a patient, or to track a patient's responsiveness to therapeutic intervention.

Description

[0001] A method of diagnosing and / or differentiating a neurological disorder Field of the invention The present invention relates to a method for diagnosing a patient having a neurological (including neuromuscular) disorder, which method extends to the pre-symptomatic diagnosis of said patients. The invention further relates to a method for distinguishing between neurological disorders, said method extending to include the ability to discriminate one subtype from other subtypes of the same disorder. Examples of the latter include the ability to discriminate: spinal muscular dystrophy type II (astasia) from spinal muscular atrophy type II or type IV (mobile but may subsequently lose ambulation); “fast- progression” amyotrophic lateral sclerosis from “slow-progression” amyotrophic lateral sclerosis; and “severe disease stage” amyotrophic lateral sclerosis from “mild disease stage” amyotrophic lateral sclerosis. In this way, the present invention allows one to monitor disease state progression of a neurological disorder in a patient, or to track a patient’s responsiveness to therapeutic intervention. Background to the invention Many neurological disorders, in particular motor neuron disorders such as spinal-bulbar muscular atrophy (SBMA), spinal muscular atrophy (SMA) and neuromuscular disorders such as facioscapulohumeral muscular dystrophy (FSHD) have known genetic causes, making them potentially amenable to genetic intervention. The progress of gene therapy opportunities into the clinic has, however, been slow and with only limited success, no doubt adding to the existing finance and resource burden pharmaceutical companies have to balance. One complicating factor in the case of clinical trials relating to neurological disorders remains an inability to identify accurately subjects having one neurological disorder from cohorts of subjects having a different neurological disorder but who present with the same or very similar symptoms. Consequently, when recruiting subjects for the purpose of a clinical trial, this cohort will inherently include subjects not meeting the correct selection criteria for the trial and giving rise to false non-responders and / or false responders. On the other hand, some neurological conditions, such as inclusion body myositis (IBM) and myasthenia gravis (MG) are at most poorly understood in terms of mechanisms and genetic contributions and remain difficult to diagnose or differentiate because of this. IBM is often misdiagnosed and symptoms can be dismissed as signs of ageing. In MG, the serum level of acetylcholine receptor antibodies, being elevated in around 80- 85% of patients, is useful for diagnosis but does not predict the disease severity nor the response to treatment. Amyotrophic lateral sclerosis (ALS) is the most common form of MND. ALS is a fatal condition resulting in progressive degeneration and death of upper and lower motor neurons, protein aggregation, severe muscle atrophy, and respiratory insufficiency. Median survival is between 2 and 5 years from the onset of symptoms. ALS can manifest as either familial ALS (fALS) (~10% of cases) or sporadic ALS (sALS) (~90% of cases). Part of the challenge to diagnosis is that several other diseases mimic ALS, especially in the early stages of disease. Among these are SBMA and SMA. SBMA is a slowly progressive disorder characterised by weakness and atrophy of facial, bulbar, and limb muscles without upper motor neuron involvement. SBMA itself is difficult to diagnose, and the disorder with which it is most often confused is ALS. It is estimated that 1 in 25 individuals diagnosed with ALS has SBMA instead of ALS. SMA is a lower MND with predominantly proximal muscle weakness and reduced or absent reflexes, in which patients are classified in 4 groups (SMA I – IV) according to their severity, and for which improved diagnostic and stratification tools are needed, especially due to the variable response of patients to the recently available treatment with Nusinersen (all-P-ambo-2'-O-(2-Methoxyethyl)-5-methyl-P-thiouridylyl-(3'→5')-2'-O-(2- methoxyethyl)-5-methyl-P-thiocytidylyl-(3'→5')-2'-O-(2-methoxyethyl)-P-thioadenylyl- (3'→5')-2'-O-(2-methoxyethyl)-5-methyl-P-thiocytidylyl-(3'→5')-2'-O-(2-methoxyethyl)-5- methyl-P-thiouridylyl-(3'→5')-2'-O-(2-methoxyethyl)-5-methyl-P-thiouridylyl-(3'→5')-2'-O-(2- methoxyethyl)-5-methyl-P-thiouridylyl-(3'→5')-2'-O-(2-methoxyethyl)-5-methyl-P- thiocytidylyl-(3'→5')-2'-O-(2-methoxyethyl)-P-thioadenylyl-(3'→5')-2'-O-(2-methoxyethyl)-5- methyl-P-thiouridylyl-(3'→5')-2'-O-(2-methoxyethyl)-P-thioadenylyl-(3'→5')-2'-O-(2- methoxyethyl)-P-thioadenylyl-(3'→5')-2'-O-(2-methoxyethyl)-5-methyl-P-thiouridylyl- (3'→5')-2'-O-(2-methoxyethyl)-P-thioguanylyl-(3'→5')-2'-O-(2-methoxyethyl)-5-methyl-P- thiocytidylyl-(3'→5')-2'-O-(2-methoxyethyl)-5-methyl-P-thiouridylyl-(3'→5')-2'-O-(2- methoxyethyl)-P-thioguanylyl-(3'→5')-2'-O-(2-methoxyethyl)guanosine). Adult forms of SMA can be difficult to distinguish from SBMA and from ALS. There is currently no definitive diagnostic test for many of the neurological and neuromuscular disorders, and patients often need to undergo multiple investigations including neuroimaging, lumbar punctures, muscle biopsies, nerve conduction studies, and / or electromyography. These tests are invasive, very resource intensive, and lead to significant delays in a diagnosis for patients. The mean time from onset of symptoms to confirmation of diagnosis in a disorder such as ALS is 12–18 months, and has remained so despite 20+ years of research. Since strong preclinical evidence suggests that effective therapy must be delivered as early as possible to slow disease progression, definitive diagnostic tests to diagnose neurological and neuromuscular disorders more quickly, and ideally at the pre-clinical level, are thus urgently needed. The current Gold Standard biomarker for diagnosis of a neurological disorder (in particular, ALS and SMA) is neurofilament protein. This class of structural proteins is expressed exclusively in neurons at high levels and, whilst their precise function remains unknown, they are thought to be critical for radial growth and stability of axons. Several mutations identified in the NEFL, NEFM and NEFH genes (encoding the neurofilament light, medium and heavy proteins, respectively) lead to abnormal neurofilament aggregation and are known to be associated with ALS and familial Parkinson’s disease (fPD). Under normal conditions, neurofilaments form highly stable structures within axons within which their turnover is low. Conversely, under diseased state conditions, these proteins accumulate, taking on aberrant subunit stoichiometry, and form a liquid crystal gel network within the axon. A loss of axon integrity follows and neurofilament proteins are released into the surrounding cerebrospinal fluid. The true biomarker potential of neurofilament protein as a diagnostic for neurological disorders has been hampered due to the relatively invasive nature of the lumbar puncture procedure by which cerebrospinal fluid samples are obtained. Moreover, where neurofilament biomarker data have been reported historically, these data are typically limited to only a handful of specific neurological disorders for which diagnostic lumbar puncture is indicated. Notwithstanding the above-noted short comings, neurofilament protein remains the current Gold Standard biomarker for the diagnosis of neurodegenerative diseases. Better biomarkers are needed for prognosis and diagnosis of these conditions. There is a general need for easy-to-access biomarkers that can circumvent challenges with respect to stability and reliability both of performance metrics and across laboratories. Summary of the invention The present invention solves one or more of said problems or shortfalls by providing a method for the predictive diagnosis of a subject having a(ny) neurological disorder. The methodology is highly robust having been validated by extensive cross-referencing to national neurological deposit samples obtained from subjects known to have a defined neurological disorder and spanning all of the more prevalent disorder types. In more detail, the present invention flows from the quantitative and qualitative analysis of SM expression levels in Extracellular Vesicles (EVs). To the surprise of the present inventors, there exists a strong correlation that links together the differential expression level of one or more SM observation(s) and the neurological disorder “status” of a subject. This correlation underpins the predictive diagnostic method of the present invention and therefore provides a robust means for confirming whether a subject is healthy or whether the subject has a defined neurological disorder (including a defined disorder subtype, and / or a defined disease-progression stage thereof). The diagnostic accuracy of the present invention may be readily dialled-up by increasing (or readily dialled down by decreasing) the number of SM observations relied on. According to a first aspect of the present invention, there is provided a method for diagnosing a first neurological disorder in a subject, said method comprising: a) providing a patient primary sphingomyelin (SM) profile for said subject: i. wherein said patient primary SM profile is obtainable by mass spectrometry analysis of a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject, and ii. wherein said patient primary SM profile comprises one or more primary SM observation(s), each of which is a different SM observation(s); b) providing a reference primary SM profile comprising one or more primary SM observations: i. wherein said reference primary SM profile is obtainable by comparative mass spectrometry analysis of body fluid samples comprising 30-300nm EVs isolated from two different reference cohorts selected from a healthy subject cohort and a cohort of subjects having the first neurological disorder; ii. wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD) value for said one or more primary SM observation(s); and (b) between each reference cohort, identification of one or more key SM observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; iii. wherein said reference primary SM profile comprising one or more of said key SM observation(s) is provided in a form that is suitable for comparative analysis with the patient primary SM profile; c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile such that: i. when the healthy subject cohort is selected as the reference primary SM profile and the mean-fold expression value(s) for said one or more key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for said one or more patient SM observation(s) falls within 2 SD of the mean expression value(s) of said corresponding one or more key SM observation(s) of the reference SM profile, said “match” being indicative the subject is healthy and / or a “non-match” being indicative the subject has the first neurological disorder; and / or ii. when the cohort of subjects having the first neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for said one or more key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for said one or more patient SM observation(s) falls within 2 SD of the mean expression value(s) of said corresponding one or more key SM observation(s) of the reference SM profile, said “match” being indicative the subject has the first neurological disorder and / or a “non-match” being indicative the subject is healthy In one embodiment, said method includes one or more corresponding method(s) run in parallel for diagnosing a second (or subsequent) neurological disorder in the subject. Each of said parallel methods may be run sequentially or simultaneously, preferably simultaneously. A second aspect of the invention provides a method for diagnosing a first or second neurological disorder, said method comprising: a) providing a patient primary sphingomyelin (SM) profile for said subject: i. wherein said patient primary SM profile is obtainable by mass spectrometry analysis of a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject, and ii. wherein said patient primary SM profile comprises one or more primary SM observation(s), each of which is a different SM observation; b) providing a reference primary SM profile comprising one or more primary SM observations: i. wherein said reference primary SM profile is obtainable by comparative mass spectrometry analysis of body fluid samples comprising 30-300nm EVs isolated from two different reference cohorts selected from a a cohort of subjects having the first neurological disorder and a cohort of subjects having the second neurological disorder; ii. wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD) value for said one or more primary SM observation(s); and (b) between each reference cohort, identification of one or more key SM observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; iii. wherein said reference primary SM profile comprising one or more of said key SM observation(s) is provided in a form that is suitable for comparative analysis with the patient primary SM profile; c) comparing said one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile such that: i. when the cohort of subjects having the first neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) falls within 2 SD of the mean expression value(s) of said corresponding one of more key SM observation(s) of the reference SM profile, said “match” being indicative the subject has the first neurological disorder and / or a “non-match” being indicative the subject has the second neurological disorder: and / or ii. when the cohort of subjects having the second neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) falls within 2 SD of the mean expression value(s) of said corresponding one of more key SM observation(s) of the reference SM profile, said “match” being indicative the subject has the second neurological disorder and / or a “non-match” being indicative the subject has the first neurological disorder, This method allows one to distinguish between two different types (including two different stages) of the same or a similar neurological disorder. In practice, this method will usually be performed on subjects who have been prior diagnosed (e.g. by this invention) as having a neurological disorder, or on subjects suspected to have a neurological disorder (e.g. a familial form of a disorder, or based on symptomatic presentation). By way of example, this aspect of the present invention allows one to distinguish between the three different subtypes of ALS (as set forth in claims 93-98), namely: ALS (no-PMA / PLS) versus PLS; ALS (no PMA / PLS) versus PMA; and PLS versus PMA. The present invention may include the introduction of additional (e.g. secondary and / or subsequent) “reference” SM profile comparisons representative of one or more further cohorts of neurological disorder(s). For example, whilst an initial “reference” SM profile comparison (e.g. healthy subject cohort versus disorder X cohort, or disorder X cohort versus disorder Y cohort) may well point towards a confirmed diagnosis of disorder X, there may be informed basis (e.g. familial knowledge, presented symptoms, and / or improved scientific understanding) to question said initial diagnosis and to seek further clarification as to whether the specific disorder in question is disorder X’ or X’’. In such a scenario, the introduction of a “reference secondary SM profile" representative of disorder X’ and / or disorder X’’ would be advantageous and provide valuable additional information. In one embodiment, this may take the form of a comparison of “reference” SM profiles for a cohort of subjects having disorder X’ versus a cohort of subjects having disorder X’’. Alternative embodiments may, for example, include cohort comparisons of healthy versus disorder X’ and healthy versus disorder X’’ (see claims 2 & 4). A third aspect of the invention provides a rapid triage screen to determine whether a subject is healthy or has a (generic) neurological disorder selected from: ALS, SBMA, SMA or PD (as set forth in claims 86-89); or ALS, SBMA, SMA, PD or AD (as set forth in claims 90-92). A fourth aspect of the invention provides a method for early-stage diagnosis of a subject having a first neurological disorder and wherein said subject presents as asymptomatic for said neurological disorder. Said method may employ any method of the invention as herein defined (see claim 99). According to a fifth aspect, the present invention provides a method for screening a subject to ensure said subject meets the correct neurological disorder qualifying criteria for enrolment into a clinical trial, said method comprising any method of the present invention as herein described. The second aspect of the present invention provides a preferred embodiment of this aspect, wherein the first neurological disorder and the neurological disorder relevant to the clinical trial are the same neurological disorder, and wherein said method confirms said subject does not have the second neurological disorder (see claims 100-101). A sixth aspect of the present invention provides a method for monitoring the responsiveness of a subject having a first neurological to therapeutic intervention, as for example is illustrated by reference to claims 102-109 (see Example 16 and Figure 10). A seventh aspect of the invention is directed to a method for monitoring disease progression, for example as set forth by reference to claims 110-112. Detailed description of the invention The present invention provides a robust method for the diagnosis of a subject having a neurological disorder, a non-exhaustive list of which includes Amyotrophic Lateral Sclerosis (ALS), familial ALS (fALS), sporadic ALS (sALS), Spinal-Bulbar Muscular Atrophy (SBMA) and Spinal Muscular Atrophy (SMA), Spinal Muscular Atrophy type II (SMAII), Spinal Muscular Atrophy type III / IV (SMAIII / IV), Parkinson’s disease (PD), Alzheimer’s disease (AD), Charcot-Marie-Tooth (CMT), Facioscapulohumeral muscular dystrophy (FSHD), Myasthenia Gravis (MG), Progressive Muscular Atrophy (PMA), Fast progression ALS, Slow progression ALS, Severe disease ALS, Mild disease ALS, Huntington disease, Duchenne muscular dystrophy, Frontotemporal Dementia (FTD), and the disease spectrum known as ALS-FTD (in which FTD clinically overlaps with ALS). In its simplest form, the present invention provides a method that compares the SM expression profile from a subject to be tested (hereinafter referred to as the “patient SM profile”) with an SM profile that defines one of two possible outcomes of the test (hereinafter referred to as the “reference SM profile”). This is illustrated by reference to claims 1 & 3, each of which defines two principal diagnostic outcomes, namely: - the subject is either healthy or has a first neurological disorder (see claim 1); or - the subject has either a first neurological disorder or a second neurological disorder (see claim 3). Naturally, either or both of these methods may form part of a series of parallel methods in which a different neurological disorder may be introduced, thereby providing the ability to screen (simultaneously or sequentially) for any bespoke combination of neurological disorders. Each reference SM profile comprises one or more SM observation(s) and a comparison of the same one or more SM observation(s) when shared in common between two reference SM profiles allows one to select one or more key SM observation(s) that demonstrates a statistically significant difference, thereby providing statistical basis upon which the two reference SM profiles are distinguishable. The patient SM profile is then compared with either (or both) reference SM profile(s) and the relative expression level for said one or more key SM observation(s) interrogated. Finally, since the observed expression level for each key SM observation(s) of the patient SM profile will match more closely to the calculated mean expression level for one reference SM profile (than it will to the other reference SM profile), the degree of said match will provide statistical basis for assigning a probability for diagnosis. The SM may be any sphingomyelin. For example, the SM fatty acid base component may be selected from any of the following fatty acid base structures: o SM base d16:1: o SM base d17:1; o SM base d18:1; o SM base d19:1; o SM base d20:1; o SM base d43:1; o SM base d34:1; o SM base d32:0; o SM base d38:3; o SM base d33:0; o SM base d38:2; o SM base d33:1; o SM base d32:1; o SM base d44:1; o SM base d42:1; o SM base d41:1; o SM base d35:1; o SM base d36:1; o SM base d44:2; o SM base d34:2; or o SM base d36:3. More particularly, the SM fatty acid base component may be selected from any of the following fatty acid base structures: o SM base d16:1: o SM base d17:1; o SM base d18:1; o SM base d19:1; or o SM base d20:1. Similarly, the SM component may be selected from: SM14:0, SM14:1, SM15:0, SM16:0, SM16:1, SM17:0, SM18:0, SM18:1, SM19:0, SM20:0, SM20:1, SM20:2, SM21:0, SM22:0, SM22:1, SM23:0, SM23:1, SM24:0, SM24:1, SM24:2, SM25:0, SM25:1, SM26:0 and SM26:1. More particularly, the SM component may be selected from: SM14:0, SM15:0, SM16:0, SM16:1, SM17:0, SM18:0, SM18:1, SM19:0, SM20:0, SM20:1, SM21:0, SM22:0, SM22:1, SM23:0, SM23:1, SM24:0, SM24:1, SM24:2, SM25:0, SM25:1, SM26:0 and SM26:1. The SM chemical structures listed above are considered the conventional SM isomer forms. In addition, naturally occurring variant forms exit, such as desaturation, hydroxylation, and methyl branching variants. For the purpose of the present invention, an SM variant form may be considered a unique SM structure in its own right or subsumed as a conventional isomer (according to its structural “best match”). Thus, to the extent an SM variant is identified, it is preferable that the SM variant be treated consistently (i.e. as a unique SM, or a conventional isomer) when preparing all SM profiles (i.e. both SM reference profiles and SM patient profile) for use in the same diagnostic method of the present invention, For reasons that relate purely to laboratory convenience (i.e. availability of a suitable mass spectrometry platform) the present inventors have focussed on the SM base d18:1 class of SMs to demonstrate the utility of the present invention. Thus, for this reason only, the base d18:1 class of SMs represents a preferred SM class of the present invention – exemplary chemical structures for this SM class are illustrated in Figure 9. For the very same convenience / illustrative purposes, in one embodiment the SM component is selected from one or more of SM15:0, SM16:0, SM16:1, SM17:0, SM18:0, SM18:1, SM19:0, SM20:0, SM20:1, SM21:0, SM22:0, SM22:1, SM23:0, SM23:1, SM24:0, SM24:1. Thus, for the avoidance of any doubt, reference to SM base d18:1 is purely illustrative of the broader genus of SM base classes that may be employed in accordance with the present invention and for which corresponding technical effects are therefore demonstrable. The present invention relies on analysis of the sphingomyelin (SM) levels present in extracellular vesicles (EVs) that have been isolated from a body sample of a test subject. Thus, throughout this disclosure, the term “a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject” may preferably mean “a sample comprising 30-300nm extracellular vesicles (EVs) isolated from a body sample of a subject”. One surprising observation noted by the inventors was that this sub-population of SMs is (highly) differentially expressed, on the one hand, between healthy subjects and subjects having a neurological disorder and, on the other hand, between subjects having different neurological disorders (including different sub-types or disease-stages of the same disorder). It was also surprising to note that SM expression profiles for the EV- isolated sub-population of SMs were significantly more robust than any corresponding SM profiles obtained from total serum, total plasma or total saliva (or poorly refined fractions thereof). In this regard, EV-isolated SMs routinely demonstrate a p-value of < 2.86E-07 whereas total plasma / serum derived SMs typically demonstrate (at best) a p-value of < 1.82E-05). One further unexpected observation was that differential expression profiles reported for the sub-population of EV-isolated SMs were consistently greater than the corresponding profiles reported for SM derived from total serum / plasma. Thus, the present inventors have developed a powerful (e.g. highly selective and / or robust) diagnostic tool that outperforms existing biomarker assays, notably assays based on “neurofilament protein” (i.e. the current “gold standard”). This is confirmed by way of the comparative data study presented as Example 12. When considering which SM observation(s) to select as a key SM observation(s) for use in a diagnostic assay of the present invention, one would typically start by identifying candidate SM observation(s) that demonstrate at least (or above) a minimum threshold criterion. In this regard, minimum threshold fold-change parameters of the present invention typically start from a minimum fold-change value of at least (or greater than) +1.1 or at least (or less than) -1.1 through to a fold-change value of at least (or greater than) +2 or at least (or less than) -2. Alternatively, when selecting the key SM observation(s) required to create a reference SM profile, one may prefer to select the SM observation that demonstrates the highest + / - fold-change value, and then proceed to select further key SM observations according to the next highest + / - fold-change and so on and so forth. This approach may be followed to select a “reference SM profile” comprising one SM observation (the highest fold-change SM observation), two SM observations (the two highest fold-change SM observations), three SM observations (the three highest fold-change SM observations), four SM observations (the four highest fold- change SM observations), and optionally five SM observations (the five highest fold- change SM observations). However, it is also necessary to ensure that adequate diagnostic robustness is keyed-in to the assay, which in turn requires that one must have confidence that the underlying cohort data are reliable as it is these data that feed directly into the “reference SM profile” selection criteria. Accordingly, each reference cohort includes (for each SM observation) a statistical mean SM expression value. This mean cohort SM expression value is compared with a corresponding mean cohort expression value for the same SM observation in the second reference disorder, and the two mean values assessed by statistical analysis. T-test and p-value are statistical concepts used to compare the means of two groups and measure the evidence against a null hypothesis. A p-value from a T-test is the probability that the results from sample data did not occur by chance, and a low p-value indicates the data did not occur by chance. So, applying the above principles to the present invention, the mean expression value for each SM observation in a first reference profile (for a first neurological disorder) is compared with the corresponding mean expression level for the same SM observation in a second reference profile (for a different neurological disorder or a healthy control). The resulting “fold-change” value and “p-value” for each of said SM comparisons.is provided in Figure 12. Next, one applies an appropriate p-value qualification criterion. Whilst any suitable p-value may be employed for cut-off purposes, for example p <0.5, p <0.25, p <0.1, p <0.075, p <0.05, p <0.025 through to p <0.005 , a p-value of p <0.05 is preferred. and includes all SM observations having a p-value that is less than 0.05, while excluding SM observations having a p-value of 0.05 or more. Thus, when a p-value of p >0.05 is superimposed on to the dataset presented in Figure 12, Figure 13 reveals those key SM observations that are available for selection according to this criterion (for each diagnostic assays considered). Having decided on an appropriate p-value (e.g. p <0.05 in Figure 13) for the diagnostic method in question, one then applies an appropriate threshold criterion in order to provide a final selection of key SM observations, any number of which may be chosen for inclusion in a “reference SM profile” of choice (for each of the considered diagnostic assays). As briefly discussed earlier, one may choose to select say the top two, three or four or five SM observations demonstrating the greatest fold-change expression values. Alternatively, one may choose to apply a threshold cut-off criterion that selects those SM observation(s) demonstrating at least a relative 2-fold expression level change. This threshold approach is purely illustrative as, in practice, one would readily consider applying a fold-change threshold cut-off starting from at least (or greater or less than) + / - 1.1-fold change and upwards in increments of 0.1, so greater than or less than + / - 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0-fold change This approach is illustrated throughout the Examples, the key variables being the selected p-value and the selected threshold criterion for expression fold-change. Particular reference is made to Examples 10, 13, 14.15 & 19. In more detail, the present invention is concerned with the particular sub-population of EVs having a size distribution of 30-300 nm. These EVs are isolated from a body fluid sample, such as total serum, total plasma or total saliva, whole blood, serum, plasma, urine, or cerebrospinal fluid, with total serum or serum being a preferred source. Importantly, EVs can also be isolated from an organ or an organoid or cell culture, e.g. from myoblasts, motor neurons, hepatocytes, cardiomyocytes and so forth, which can be obtained directly from an organ biopsy or indirectly created using induced pluripotent stem cell technology. An example is provided in Fig.1. The 30-300 nm sub-population of EVs can be isolated (e.g. from total serum, total plasma, total saliva, etc) by any conventional means. One convenient method relies on conventional centrifugation techniques, comprising a first spin at say 1,500-3,000 g (eg. ~2000 g) for say 20-40 mins (e.g. 25-35 mins, or ~30 mins), followed by recovery of the supernatant, and then a second spin at 8,000-15,000 g (e.g. 9,000-13,000 g, or ~10,000 g) for say >5-20 mins (e.g. 8-15 mins, or ~10 mins). The first spin is intended to remove large molecules having a size greater than say 500 nm (e.g. greater than 300 nm, or greater than 250 nm), and the second spin is intended to remove small molecules having a size less than say 20 nm (e.g. less than 30 nm, or less than 40 nm), When preparing the EV preparation, an inert polymer reagent, such as polyethylene glycol (PEG), is typically added to the supernatant recovered from the first spin. This provides a density gradient for the second spin, through which the EV fraction travels prior to pelleting. Finally, the ERV fraction (pellet) is resuspended in a conventional buffer, such a phosphate buffer saline (PBS). EV preparations obtained this way are positive for classical markers of exosome-like EVs (e.g. different tetraspanins such as CD63 and / or CD82 and / or CD81 and / or CD9, and / or CD71, and or positive for ALIX and / or Flotillin-1, and / or Hsp70, and / or Hsp90, and / or TSG101, and / or LAMP1, and / or LAMP2, and / or Rab7, and / or Rab11). Different methods can be used to extract EVs, including: differential ultracentrifugation, polymer-based precipitation, size exclusion chromatography, ultrafiltration, and / or immunoaffinity. When constructing a reference SM profile for each diagnostic outcome (e.g. a healthy subject, or a subject having said first or second neurological disorder), the number (“n”) of subjects per cohort should be as large as possible (e.g. n >20, or n >50), thereby providing a robust calculated mean value for each SM observation). When defining key SM observations between neurological disorder cohorts and / or healthy subject cohorts, it is often convenient to express the SM relationships that one observes by mathematical means. Mathematical modelling of this type is conventional and allows one to derive a formula or a set of mathematical operations into which is then input the observations of SM levels for a new sample (isolated from a subject to be tested), and from which an output is provided in the form of a predicted probability score that the subject has or does not have the neurological disorder in question. Comparative statistical and / or mathematical analysis of a patient SM profile and one or more reference SM profile(s) provides for accurate and automated diagnosis. An example of this approach is the use of logistic regression modelling (e.g. binomial model) to predict whether a subject SM profile finds closest nexus to one of two different diagnostic outcomes. This type of approach is therefore suitable for determining whether a patient SM profile (obtained from a subject to be tested) has closest nexus to a reference SM profile from a cohort of subjects having a first neurological disorder or to a reference SM profile from a cohort of healthy subjects (or whether said patient SM profile has closest nexus to a reference SM profile from a cohort of subjects having a first neurological disorder or to a reference SM profile from a cohort of subjects having a second neurological disorder, and thereby providing a probability score based diagnosis that the subject (patient) either has or does not have said first neurological disorder (or that the subject either has the first or second neurological disorder). This is illustrated by way of the following formula: where P (y = | x) is the probability of the sample belonging to class i given the values of its features x1, x2, x3, up to the number of features p, and β0, β1, β2, and β3are the intercept and coefficients, respectively, up to the number of coefficients (also p, there being one coefficient per feature). Working examples of a binomial modelling approach are illustrated in: Example 3 for accurate diagnosis of MND (ALS or SBMA) subjects versus healthy subjects, and for accurate diagnosis of ALS subjects versus SBMA subjects; Examples 8-11 for accurate diagnosis of PD subjects versus heathy subjects; Example 12 for accurate diagnosis of ALS subjects versus healthy subjects, and for healthy subjects versus subjects having any one of a defined group of neurological disorders; Example 13 for NMD subjects versus healthy subjects; Example 14 for ALS subtyping of ALS subjects; Example 15 for SMA subtyping of SMA subjects; and Example 16 for monitoring responsiveness to therapeutic intervention. Each of these Examples illustrates the diagnostic power of the present invention. For example, use of a single SM observation is very capable of demonstrating a diagnostic accuracy of greater than 70%. Moreover, said diagnostic accuracy may be readily dialled- up to 80-100% accuracy by the simple inclusion of one or more further SM observations, In more detail, Example 9 confirms a diagnostic accuracy of 97% was achieved with a single SM observation (e.g. C21:0. C23:1, or C24:0), which could be readily dialled-up by inclusion of a second SM observation to 98.6% (e.g. C21:0 & C24:0) or to 100% (e.g. C21:0 & C23:1, or C23:1 & C24:0). Taking this approach one step further, comparative statistical and / or mathematical analysis may be readily extended to compare three or more different diagnostic outcomes. An example of this approach is the use of logistic regression modelling (e.g. multinomial model) to predict whether a subject (patient) SM profile finds closest nexus to one of three (or more) different reference SM profiles, each being representative of a different cohort of subjects (e.g. a healthy subject cohort, a cohort of subjects having a first neurological disorder, and a cohort of subjects having a second neurological disorder). This is illustrated by way of reference to the formula: where P (y = | x) is the probability of the sample belonging to class i given the values of its features x1, x2, x3, up to the number of features p, and β0i, β1i, β2i, and β3iare the intercept and coefficients, respectively, for class i, up to the number of coefficients (also p, there being one coefficient per feature). The denominator of the equation is the sum of the exponentiated linear predictors for all the classes from j to the total number of classes k. Working examples of this approach are illustrated in: Examples 2-6 for accurate diagnosis of ALS subjects versus SBMA subjects versus healthy subjects; Example 12 for accurate diagnosis of ALS subjects versus healthy subjects versus subjects having any one of a defined group of neurological disorders (selected from ALS, PD, SBMA or SMA) versus healthy subjects; and Example 13 extends the defined group of neurological disorders in Example 12 to include AD subjects. Each of these Examples illustrates the diagnostic power of the present invention. For example, Table 7 (see Example 10) confirms that a diagnostic accuracy of 95% was achieved with a panel of 2 SM observations (C15:0 & C21:0) – see Set E, and that this level of diagnostic accuracy could be readily dialled-up to 96% by the addition of one further SM observation (C22:0) – see Set D. By including 5 further SM observations (see Set B = 8 observations in total) this accuracy was readily dialled-up to 99%. Table 7 (see Example 10) separately reports a diagnostic accuracy of 92% for a panel of 7 SM observations (see Set C), which was readily dialled-up to 99% by the addition of one further SM observation (C21:0) – see Set C*. The present invention is now described by reference to particular diagnostic embodiments in which a patient SM profile (obtained from a subject to be tested) is compared with the reference SM profile for a healthy subject and / or with a reference SM profile for subject having a first neurological disorder. Predictive diagnosis is determined according to whether said patient SM profile demonstrates closest nexus to the reference SM profile of the healthy subject or to the reference profile for a subject having the first neurological disorder. In one embodiment, the method provides for diagnosis of a subject having Alzheimer’s disease (AD), wherein the reference SM profiles compared are AD versus healthy, and wherein the patient and reference SM profile(s) share in common one or more key SM observation selected from: C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C23:1, C24:2, C24:1, C24:0; preferably from: C15:0, C18:1, C21:0, C24:2, C24:1; and more preferably from C15:0, C24.2, optionally C24.1. In a preferred embodiment thereof, the “patient” and “reference” SM profile(s) share in common at least four or five SM observations selected from: C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C23:1, C24:2, C24:1, C24:0; preferably from: C15:0, C18:1, C24:2, C24:1, optionally C21:0. In one embodiment, the method provides for diagnosis of a subject having monogenic or sporadic amyotrophic lateral sclerosis (ALS), wherein the reference SM profiles compared are ALS versus healthy, and wherein the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15:0, C16:1, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22.0, C23:1, C23:0, C24:2, C24:1, and C24:0; or C15:0, C16:0, C20:1, C20:0, C21:0, C22:1, C22:0, C24:1, and C24:0; preferably from: C15:0, C16:0, C20:1, C21:0, C22:1, C22.0, C23:0, C24:0; and more preferably from C15, C21:0, optionally C23:0. In a preferred embodiment thereof, the “patient” and “reference” SM profile(s) share in common at least four or five SM observations selected from: C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C23:1, C24:2, C24:1, C24:0; or C15:0, C16:0, C20:1, C20:0, C21:0, C22:1, C22:0, C24:1, and C24:0; preferably from: C15:0, C16:0, C20:1, C21:0, C22:1, C22.0, C23:0, C24:0; and more preferably from: C15:0, C21:0, C23:0, C22:0, optionally C22:1. In one embodiment, the method provides for diagnosis of a subject having myasthenia gravis (MG), wherein the reference SM profiles compared are MG versus healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15:0, C16:1, C16:0, C18:1, C20:1, C21:0, C22:1, C23:1, C23:0, C24:2, C24:1, C24:0; preferably from: C15:0, C16:1, C16:0, C18:1, C20:1, C21:0, C23:1, C23:0, C24:2, C24:0; and more preferably from: C21:0, C23:0, optionally C23.1. In a preferred embodiment thereof, the “patient” and “reference” SM profile(s) share in common at least four or five SM observations selected from: C15:0, C16:1, C16:0, C18:1, C20:1, C21:0, C22:1, C23:1, C23:0, C24:2, C24:1, C24:0; preferably from: C15:0, C16:1, C16:0, C18:1, C20:1, C21:0, C23:1, C23:0, C24:2, C24:0; and more preferably from C21:0, C23:0, C23.1, C15:0, optionally C21:0. In one embodiment, the method provides for diagnosis of a subject having Parkinson’s disease (PD), wherein the reference SM profiles compared are PD versus healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C16:1, C16:0, C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0; or C16:0, C20:0, C20:1, C21:0, C22:0, C23:0, C23:1, C24:0, and C24:2. preferably from: C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:0; and more preferably from: C23:1, C21:0, optionally C20:1. In a preferred embodiment thereof, the “patient” and “reference” SM profile(s) share in common at least four or five SM observations selected from: C16:1, C16:0, C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0; or C16:0, C20:0, C20:1, C21:0, C22:0, C23:0, C23:1, C24:0, and C24:2. preferably from: C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:0; and more preferably from C23:1, C21:0, c20:1, C20:0, optionally C24:0. In one embodiment, the method provides for diagnosis of a subject having Progressive muscular atrophy (PMA), wherein the reference SM profiles compared are PMA versus healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15:0, C16:0, C20:0, C21:0, C22:0, C23:0, C24:0; and preferably from: C21:0, C23:0, optionally C15:0. In a preferred embodiment thereof, the “patient” and “reference” SM profile(s) share in common at least four or five SM observations selected from: C15:0, C16:0, C20:0, C21:0, C22:0, C23:0, C24:0; and preferably from: C21:0, C23:0, C15:0, C22:0, optionally C16:0. In one embodiment, the method provides for diagnosis of a subject having Spinal and bulbar muscular atrophy (SBMA), wherein the reference SM profiles compared are SBMA versus healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from C15:0, C16:1, C16:0, C18:1, C18:0, C20:1, C20:0, C22:1, C23:0, C24.2, C24.1, C24:0; or C15:0, C16:0, C16:1, C18:0, C18:1, C20:0, C20:1, C22:0, C22:1, C23:1, C24:0, C24:1, and C24:2; or C15:0, C16:0, C18:0, C20:1, C20:0, C22:1, C22:0, C23:1, C23:0, C24:2, and C24:1; preferably from: C15:0, C22:1, C20:1, C24.2, C16:0, C23.1, C24.1; and more preferably from: C15:0, C22:1, optionally C20:1. In a preferred embodiment thereof, the “patient” and “reference” SM profile(s) share in common at least four or five SM observations selected from: C15:0, C16:1, C16:0, C18:1, C18:0, C20:1, C20:0, C22:1, C23:0, C24.2, C24.1, C24:0; or C15:0, C16:0, C16:1, C18:0, C18:1, C20:0, C20:1, C22:0, C22:1, C23:1, C24:0, C24:1, C24:2; or C15:0, C16:0, C18:0, C20:1, C20:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1; preferably from: C15:0, C22:1, C20:1, C24.2, C16:0, C23.1, C24.1; and more preferably: C15:0, C22:1, C20:1, C24.2, optionally C16:0 or C23.1. In one embodiment, the method provides for diagnosis of a subject having Spinal muscular atrophy (SMA), wherein the reference SM profiles compared are SMA versus healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24.2, C24.1, C24:0; or C15:0, C16:0, C18:0, C18:1, C20:1, C22:0, C22:1, C24:0, and C24:2; preferably from: C15:0, C22:1, C20:1, C16:0; and more preferably from: C15:0, C22:1, optionally C20:1. In a preferred embodiment thereof, the “patient” and “reference” SM profile(s) share in common at least four or five SM observations selected from: C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24.2, C24.1, C24:0; or C15:0, C16:0, C18:0, C18:1, C20:1, C22:0, C22:1, C24:0, and C24:2. preferably from: C15:0, C22:1, C20:1, C16:0, optionally C18:1. The present invention is now described by reference to particular diagnostic embodiments in which a patient SM profile (obtained from a subject to be tested) is compared with a reference SM profile for a subject having a first neurological disorder and with a reference SM profile for subject having a second neurological disorder. Predictive diagnosis is determined according to whether said patient SM profile demonstrates closest nexus to the reference SM profile for the subject having the first neurological disorder or to the reference profile for a subject having the second neurological disorder. In one embodiment, the method provides a screen for distinguishing a subject having Fast progression ALS from a subject having Slow progression ALS, wherein the reference SM profiles compared are Fast versus Slow ALS, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 22-23. In one embodiment, the method provides a screen for distinguishing a subject having Severe ALS from a subject having Mild ALS, wherein the reference SM profiles compared are Severe versus Mild ALS, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 24-25. In one embodiment, the method provides a screen for distinguishing a subject having AD from a subject having ALS, wherein the reference SM profiles compared are AD versus ALS, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 26-27.: In one embodiment, the method provides a screen for distinguishing a subject having AD from a subject having MG, wherein the reference profiles compared are AD versus MG, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 28-29.: In one embodiment, the method provides a screen for distinguishing a subject having AD from a subject having PD, wherein the reference profiles compared are AD versus PD, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 30-31. In one embodiment, the method provides a screen for distinguishing a subject having AD from a subject having PMA, wherein the reference profiles compared are AD versus PMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 32-33. In one embodiment, the method provides a screen for distinguishing a subject having AD from a subject having SBMA, wherein the reference profiles compared are AD versus SBMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 34-35. In one embodiment, the method provides a screen for distinguishing a subject having AD from a subject having SMA, wherein the reference profiles compared are AD versus SMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 36-37.: In one embodiment, the method provides a screen for distinguishing a subject having ALS from a subject having MG, wherein the reference profiles compared are ALS versus MG, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 38-39. In one embodiment, the method provides a screen for distinguishing a subject having ALS from a subject having PD, wherein the reference profiles compared are ALS versus PD, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 40-41.: In one embodiment, the method provides a screen for distinguishing a subject having ALS from a subject having PMA, wherein the reference profiles compared are ALS versus PMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) as set forth in claims 42-43. In one embodiment, the method provides a screen for distinguishing a subject having ALS from a subject having SBMA, wherein the reference profiles compared are ALS versus SBMA, and the “patient” and “reference” SBMA profile(s) share in common one or more SM observation(s) as set forth in clams 44-45. In one embodiment, the method provides a screen for distinguishing a subject having ALS from a subject having SMA, wherein the reference profiles compared are ALS versus SMA, and the “patient” and “reference” SBMA profile(s) share in common one or more SM observation(s) as set forth in claims 46-47. In one embodiment, the method provides a screen for distinguishing a subject having MG from a subject having PD, wherein the reference profiles compared are MG versus PD and the “patient” and “reference” SBMA profile(s) share in common one or more SM observation(s) as set forth in claims 48-49. In one embodiment, the method provides a screen for distinguishing a subject having MG from a subject having PMA, wherein the reference profiles compared are MG versus PMA and the “patient” and “reference” SBMA profile(s) share in common one or more SM observation(s) as set forth in claims 50-51. In one embodiment, the method provides a screen for distinguishing a subject having MG from a subject having SBMA, wherein the reference profiles compared are MG versus SBMA and the “patient” and “reference” SBMA profile(s) share in common one or more SM observation(s) as set forth in claims 52-53. In one embodiment, the method provides a screen for distinguishing a subject having MG from a subject having SMA, wherein the reference profiles compared are MG versus SBMA and the “patient” and “reference” SMA profile(s) share in common one or more SM observation(s) as set forth in claims 54-55. In one embodiment, the method provides a screen for distinguishing a subject having PD from a subject having PMA, wherein the reference profiles compared are PD versus PMA and the “patient” and “reference” SMA profile(s) share in common one or more SM observation(s) as set forth in claims 56-57. In one embodiment, the method provides a screen for distinguishing a subject having PD from a subject having SBMA, wherein the reference profiles compared are PD versus SBMA and the “patient” and “reference” SMA profile(s) share in common one or more SM observation(s) as set forth in claims 58-59. In one embodiment, the method provides a screen for distinguishing a subject having PD from a subject having SMA, wherein the reference profiles compared are PD versus SMA and the “patient” and “reference” SMA profile(s) share in common one or more SM observation(s) as set forth in claims 60-61. In one embodiment, the method provides a screen for distinguishing a subject having PMA from a subject having SBMA, wherein the reference profiles compared are PMA versus SBMA and the “patient” and “reference” SMA profile(s) share in common one or more SM observation(s) as set forth in claims 62-63. In one embodiment, the method provides a screen for distinguishing a subject having PMA from a subject having SMA, wherein the reference profiles compared are PMA versus SMA and the “patient” and “reference” SMA profile(s) share in common one or more SM observation(s) as set forth in claims 64-65. In one embodiment, the method provides a screen for distinguishing a subject having SBMA from a subject having SMA, wherein the reference profiles compared are SBMA versus SMA and the “patient” and “reference” SMA profile(s) share in common one or more SM observation(s) as set forth in claims 66-67. In one embodiment, the method provides a screen for distinguishing a subject having SMAII from a subject having SMAIII / IV, wherein the reference profiles compared are SMAII versus SMAIII / IV and the “patient” and “reference” SMA profile(s) share in common one or more SM observation(s) as set forth in claims 68-69. Specific SM panels employed throughout the Examples are presented as further embodiments of the present invention as set forth in claims 60-114. One embodiment of the present invention provides for a diagnostic outcome in which a subject is diagnosed as either “healthy” or as having ALS, SBMA, SMA or PD, or as having a “neurological disorder” in the form of a Disease Control status (not further specified though one selected from SBMA, SMA, or PD as set forth in claims 86-89). Said Disease Control class may be readily extended to include further neurological disorders, such as AD. se Methods such as these provide an extremely powerful initial triage step for subjects suspected of having a neurological disorder, and acts as a pivotal gateway in patient management. In a related embodiment, the present invention provides for a diagnostic outcome in which a subject is diagnosed as having either ALS or as having a Disease Control status “neurological disorder” in the form of a Disease Control status (not further specified though one selected from SBMA, SMA, or PD). The present invention also provides a highly specific means for subtyping subjects, such as subjects having different types / forms of ALS, for example: Fast progression ALS and Slow progression ALS – as set forth in claims 22-23; Severe ALS and Mild ALS – as set forth in claims 24-25; SMA type III / IV and SMA type II – as set forth in claims 68-69; and ALS versus PMA versus PLS - as set forth in claims 93-98. Without the ability to provide this level of subtyping specificity, patients are often misdiagnosed, resulting in sub-optimal treatment and / or incorrect palliative care provisions being implemented. Delayed presentation of the symptoms of a neurological disorder is not uncommon and the present invention is able to interrogate beyond this late-stage diagnostic window. In doing so, the present invention is therefore able to provide diagnosis of a latent neurological disorder several to many years ahead of when a subject becomes symptomatic – see claim 99. Naturally, this provides a much improved therapeutic window of opportunity to address the neurological disorder and, in particular, at an earlier disease- stage when the condition may be most susceptible to therapeutic intervention. This high level of diagnostic sensitivity is important when selecting candidates for inclusion in clinical trials, and therefore provides a critical screen to ensure that all recruited candidates do in fact have the correct neurological disorder that is the subject of the clinical trial – see claims 100-101. Moreover, said diagnostic sensitivity of the present invention therefore immediately lends itself as a perfect tool for tracking responsiveness to therapeutic intervention (see, for example claims 102-109) or for monitoring disease state progression (see, for example, claims 110-113) in a subject having a neurological disorder. In this regard, the invention allows one to track disease state transitioning or reversion from one neurological disorder subtype to the next. Brief description of the drawings The present invention is described with reference to the appended non-limiting examples and the accompanying drawings, in which: Figure 1 illustrates sphingomyelin profiles of extracellular vesicles from cell culture medium and from sera, which are different between SBMA, SMA, ALS and healthy controls, wherein A is a heat map showing the sphingomyelin profiles of exosomes secreted into culture medium by ALS and healthy muscle cells (n=6 / group), wherein B represents sphingomyelin levels in extracellular vesicles in sera samples from ALS, SBMA, SMA and healthy control subjects (n=10 per group); and wherein HC: healthy control, ALS: Amyotrophy Lateral Sclerosis, SBMA: Spinal and Bulbar Muscular Atrophy, SMA: Spinal Muscular Atrophy, and SM: Sphingomyelin; Figure 2 illustrates sphingomyelin profiles of extracellular vesicles from sera, which can be used as biomarkers to distinguish SBMA, ALS and healthy subjects, wherein A is principal component analysis showing distributions of SBMA, ALS and healthy subjects on principal components 1 and 2 (x axis and y axis respectively), wherein n=116 ALS, n=62 SBMA, n=117 healthy control; wherein B-D represent a multinomial regression model using the extracellular vesicle SM profiles from SBMA, ALS and healthy sera, wherein a random 50% of the cohort was used to build the model, which correctly predicted disease status of the remaining 50%, wherein B is an ROC curve showing the performance of the classification model at all classification thresholds, wherein the area under the curve is 1, representing 100% sensitivity (no false negatives) and 100% specificity (no false positives); wherein C is a confusion matrix showing the accuracy of the model to allocate subjects to the correct group; wherein D illustrates no age effect and E illustrates no gender effect observed in SM level in ALS, healthy or SBMA groups (with C15:0 shown as an example); and wherein ALS: Amyotrophy Lateral Sclerosis, SBMA: Spinal and Bulbar Muscular Atrophy, ROC: Receiver Operating Characteristic curve, and AUC: Area Under the Curve; Figures 3 A – O illustrate levels of sphingomyelins in EVs extracted from sera of ALS, SBMA, SMA, PD, AD, or healthy subjects, wherein each data point represents a subject and for each sphingomyelin, and significant differences are indicated where they exist between one or more groups, with p-values given as ****: P<0.0001, ***: P<0.001, **: P < 0.01, and *: P<0.05; Figure 4 illustrates a nominal logistic model confusion matrix obtained from a model built using data either from 15 SMs from Figures 3A-O or from a subset of 5 SMs (C15:0, C24:2, C21:0, C22:0, and C24:1); Figures 5 A – F illustrate levels of selected sphingomyelins from extracellular vesicles in sera of different NMDs and neurodegenerative conditions, wherein ALS: Amyotrophy Lateral Sclerosis (n=116), CMT: Charcot-Marie-Tooth (n=2), FSHD: Facioscapulohumeral Muscular Dystrophy (n=1), Healthy: healthy control (n=117), MG: Myasthenia Gravis (n=4), PD: Parkinson’s disease (n=16), SBMA: Spinal and Bulbar Muscular Atrophy (n=62), SMAII: Spinal Muscular Atrophy type II (n=5), and SMA III / IV: Spinal Muscular Atrophy type III / IV (n=5); Figures 6 A – G illustrate levels of sphingomyelins from extracellular vesicles in sera of SMA type II, SMA type III / IV patients and healthy subjects, wherein ****: P<0.0001, ***: P<0.001, Healthy: healthy control (n=117), SMAII: Spinal Muscular Atrophy type II (n=6), and SMA III / IV: Spinal Muscular Atrophy type III / IV (n=5); Figure 7 illustrates no effect was observed of age (A) or sex (B) on levels of SM, with an example shown for C15:0; Figure 8 illustrates SM levels are stable (A); and consistent between sample collection sites (B-F); Figure 9 illustrates exemplary structures of the base d18:1 class of sphingomyelins; Figure 10 illustrates levels of sphingomyelins from extracellular vesicles in sera of healthy mice (wild-type; WT), and of SMA mice either untreated (-) or treated (+) with AAV9-SMN therapy. ****, ***, ** and * significantly different from untreated values with P<0.0001, P<0.001, P<0.01 and P<0.05 respectively. Sample numbers: SMA untreated n = 11; SMA treated n = 12; wild-type n = 10; Figure 11 illustrates a further embodiment of the present invention, which is based on the “decision tree” (or “selection algorithm”) approach and is described in more detail in Example 18; Figure 12 illustrates expression level fold-changes for a wide range of SM observations tested, and across a wide selection of neurological disorders. Fold change values and p- values (when n>=3) are illustrated; Figure 13 illustrates the same data are shown in Figure 12. Purely for illustrative purposes a relative p-value cut-off has been applied, on this occasion at a p-value < 0.05.

[0002] Examples Materials and Methods EV extraction: EV extraction from cell culture Myoblasts were grown in 20% foetal bovine serum (FBS) proliferation medium with Dulbecco’s modified Eagle’s medium (DMEM) / M199 4X (Thermofisher) mix, supplemented with 5ng / ml Epidermal growth factor, 0.5ng / ml basic fibroblast growth factor (bFGF), 25 mg / ml fetuin and 5 mg / ml insulin. The purity of the culture was assessed by 80 to 90% Desmin positive immunostaining and CD56 sorting. When reaching 4.6 million, cells were plated and differentiated the following day in DMEM (FBS and supplement-free medium). The remaining proliferating media was previously washed-off by five washes with DMEM. Muscle cells were differentiated for three days before collecting the media for exosome extraction as previously described (https: / / pubmed.ncbi.nlm.nih.gov / 32641118 / ). Dead cells and debris were removed from the collected media by two steps of centrifugation at 260 g for 5-10 min and 2000 g at 4°C for 20 min. The supernatant was then filtered (0.22mm) before addition of the Total Exosome Isolation kit from cell culture media (Thermofisher) following the manufacturer instructions. Exosomes were then re- suspended in 70 ml of PBS for mass spectrometry analysis. EV extraction from serum samples 50 microliters (for mice) to 150 microliters (for human) of serum of ALS, SBMA, SMA, and healthy human subjects (n=10 of each) was thawed on ice and centrifuged at 2,000g for 30 min at 4°C. The supernatant was transferred to a clean tube on ice.30 microliters (or 10 microliters for 50 microliters of serum) of Total Exosome Isolation (from serum) Life Technologies® was added to the supernatant, and the sample vortexed. The sample was incubated on ice for 30 min, and then centrifuged at 10,000g for 10 min at room temperature. The supernatant was removed carefully, and the pellets were re-suspended in 50 microliters of clean PBS and stored at -80⁰C until processed for mass spectrometry analysis. Lipid analysis by mass spectrometry: Chemicals Methanol and chloroform (HPLC grade) and α-Cyano-4-hydroxycinnamic acid (CHCA) were purchased from Sigma Aldrich, UK. N-palmitoyl-d31-D- erythrosphingosylphosphorylcholine used as an internal standard (ISTD) and hexadecanoyl sphingomyelin (SM [C16:0] used as calibrant) were purchased from Avanti Polar Lipids, Inc, AL, US. Extraction of lipids Lipid extraction from extracted exosomes from both serum and cell samples was adapted from Hori et al., 2016 (http: / / dx.doi.org / 10.1016 / j.cca.2015.09.009). Briefly, 50 microliters of exosome extract from serum or cell culture was accurately added into a 1.2 ml microcentrifuge (“Eppendorf”) tube. Then, 50 microliters of internal standard (0.1 mM in 2:1 v / v chloroform: methanol) was added with further addition of 600 microliters of chloroform:methanol (2:1). The mixture was then vortexed at full power for 2 minutes. Then, 200 microliters of “Elga” ultrapure (Type I) purified water was added to the mixture and vortexed for a further 2 minutes at full power. Centrifugation was then performed for 10 minutes at 13000rpm (4⁰C) and 400 microliters was removed from the chloroform phase containing unbound lipids and transferred to a fresh 1.2ml microcentrifuge (“Eppendorf”) tube. The solvent was then evaporated to dryness under a gentle stream of nitrogen at 40⁰C and the residue was reconstituted in 50 microliters of chloroform: methanol (2:1). A 10 microliter aliquot was removed from the reconstituted sample and transferred to a fresh 1.2ml microcentrifuge (“Eppendorf”) tube containing 20 microliters of CHCA (5 mg / ml in chloroform / methanol, 2 / 1 v / v) and vortexed for 1 minute at full power. Finally, 0.5 μl of the mixture was applied to the MALDI sample plate using the forced dried-droplet method, the sample plate was then inserted into the MALDI-TOF / TOF MS analyser. Mass spectrometry analysis Mass spectrometry was performed using a MALDI-TOF / TOF MS instrument (ABSCIEX TOF / TOFTM4800 system, ABSCIEX, MS, US) in both reflector and MSMS 1Kv modes to enhance spectra resolution for accurate quantitation and SMs identification. SMs were measured (quantification mode) in reflector positive ion mode with a mass range of 600- 850 Da, a focusing mass at 700 Da and 150 laser shots per spot with a laser power of 3500 AU. For MSMS mode (identification mode), a 1Kv acceleration was used with 400 laser shots per spot at a laser power of 4500 AU. Relative precursor isolation window was 200 FWHM using metastable suppressor. All SMs precursor ions were subjected to MS / MS for identification purposes and product ion masses saved as .mgf files using T2D to Mascot converter (ABSCIEX, MA, US) for database searching. SM identification was performed using Lipid Blast version Full- Release-3 (https: / / doi.org / 10.1038 / nmeth.2551, Fiehn Lab, CA, US). SMs levels were calculated by the y=ax + b equation derived from the linear regression analysis (average of n=3 spots / sample) (1 / x weighted) of 5 calibrations SM levels at (0.01, 0.1, 1, 5, 10 mM) using intensity / concentration ratios against the ISTD using Origin Pro version 15.0 (Origin Lab, MA, US). For samples signal below the LOQ (0.01 mM) <LOQ was reported. Linearity, accuracy, and reproducibility (validation) The dynamic range of the MALDI-TOF / TOF MS SM assay was assessed between 0.01 mM-10 mM using hexadecanoyl sphingomyelin (SM [C16:0]) as standard with a fixed concentration of ISTD at 0.1 mM. Calibration curve was obtained from matrix matching calibration standards (MMCSTDs 1-5 levels / triplicate spots), with the lowest concentration standard at the LOQ (0.01mM). Reproducibility / accuracy was assessed by the evaluation of the standards deviation from the nominal concentration (back calculated concentration). All standards shown accuracy at 100±15% ranging from 88.0% to 110.0% with a variation coefficient (CV)% below 5 and a linear regression (1 / x weighted) coefficient ® >0.995. A panel of 15 sphingomyelins (SMs) was measured systematically in the samples tested: C15:0, C16:1, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0. Where numbered SM1 to SM15 in any description or drawings, the numbering corresponds to the order of the list given here. Example 1 - Cell culture findings On characterizing the sphingolipid composition of secreted exosomes from ALS cells, marked changes in the levels of a specific class of sphingolipid called the sphingomyelins (SM) were observed (Figure 1A). Sphingomyelins, originally identified as constituents of the myelin sheath, are a class of sphingolipid involved in extracellular vesicles biogenesis, membrane permeability, and inflammatory response. SMs have a specific relevance to the neuromuscular system as they are affected in ALS and SMA body fluids. To explore the potential of SMs as circulating biomarkers, vesicles extracted from the blood sera of a pilot cohort of n=10 per group of ALS, SBMA, and SMA patients were tested. This pilot showed distinct SM profiles per disease (Figure 1B). Example 2 - Ability to distinguish ALS v SBMA v Healthy As SBMA and ALS are motor neuron disorders which can be misdiagnosed with one another due to overlapping symptoms, serum samples from a larger cohort of ALS (n=116), SBMA (n=62), and Healthy (n=117) subjects were obtained. It was observed that the SM profiles of extracellular vesicles were different between SBMA v ALS v Healthy: on unbiased principal component analysis, healthy samples clustered tightly, with almost no overlap with disease samples; ALS and SBMA samples clustered more diffusely, with some overlap (Figure 2A). To test the predictive power of the SM panel, the samples were randomly split 50 / 50 into a training set and a test set. A multinomial regression model built on the training set was able to accurately assign disease status to every sample of the test set (Figure 2C). The model was blind to the disease status and to any other descriptive parameters of the test set: perfect prediction of all 148 test cases was achieved using only the levels of the 15 SM markers. No age or gender effect was observed on the SM profiles (Figures 2C & 2D). Example 3 - Predictive modelling (binomial and multinomial) using a selection of SM panels Binomial modelling was employed to provide predictive diagnosis of whether a subject is either healthy or has a MND (either ALS or SBMA). Separate binomial modelling was performed to confirm whether a subject has either ALS or SBMA. A selection of different SM panels was employed, ranging from panels comprising a single SM observation (e.g. C16:0 or C18:0) or two SM observations (e.g. C16:0 & C18:0), through to panels comprising 13 SM observations or 15 SM observations (Table 1). The 15 SM panel included SM observations: C15:0, C16:1, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0; and the 13 SM panel included SM observations: C15:0, C16:1, C18:1, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0. Binomial modelling is discussed in more detail by way of Examples 8-16. Multinomial modelling was then employed to provide predictive diagnosis of a subject, wherein the diagnosis is simultaneously selected from three or more diagnostic outcomes (in this case, healthy, ALS, or SBMA), and is discussed in more detail by way of Examples 2-6 & Example 12, In this regard, Examples 4-6 describe multinomial modelling with the above-noted 5 SM panels and four further SM panels, namely two panels comprising a single SM observation (e.g. C15:0 or C21:0), one panel comprising two SM observations (e.g. C15:0 & C21:0), and one panel comprising 11 SM observations. Example 12 illustrates multinomial modelling with three different SM panels to distinguish ALS subjects versus “generic neurological disorder” subjects versus healthy subjects, and Example 13 extends the “generic neurological disorder” to include AD subjects. In each case, the samples were randomly split 50 / 50 into a training set and a test set. An appropriate model was built on the training set, and the model was applied to the test set. The ROC-AUC score was calculated to summarize the predictive performance of the model. In every case, the panel of 15 SMs is superior to C16:0 or C18:0 (individually), or C16:0 & C18:0 (combined). The sub-panel of 13 SMs (i.e. excluding C16:0 and C18:0) is also superior in every case to C16:0 / C18:0 (individually or combined). Table 1. Results of predictive modelling for several classification tasks. Example 4 - A worked example using multinomial statistical analysis to provide predictive diagnosis of a subject having ALS versus SBMA versus healthy A multinomial regression model was created to predict the probability of a subject being Healthy or having ALS or SBMA. This was based on the differential expression measurements of a subpopulation sphingomyelins isolated from EVs in a training set of 147 subjects each having one of these three disease conditions (N = 58 Healthy; 58 ALS; 31 SBMA). From this model was derived a mathematical formula to obtain the probability of an individual having each disease condition, given as input the SM measurements in the EVs of serum from that individual. Having created such a multinomial regression model, and taking as features the EV sphingomyelin (SM) measurements of a given serum sample from an individual patient or subject, and as prediction classes the disease conditions that one wishes to distinguish (in this example, Healthy versus ALS versus SBMA), then by way of using said features to determine the class of a given sample, the following formula is used to obtain the probability that the individual has a given disease condition: where P ( y = i | x) is the probability of the sample belonging to class i given the values of its features x1, x2, x3, up to the number of features p, and β0i, β1i, β2i, and β3iare the intercept and coefficients, respectively, for class i, up to the number of coefficients (also p, there being one coefficient per feature). The denominator of the equation is the sum of the exponentiated linear predictors for all the classes from j to the total number of classes k. This Example illustrates use of an SM panel comprising 15 different SM observations (SM.C15.0, SM.C16.1, SM.C16.0, SM.C.18.1, SM.C.18.0, SM.C.20.1, SM.C20.0, SM.C21.0, SM.C22.1, SM.C22.0, SM.C23.1, SM.C23.0, SM.C24.2, SM.C24.1, SM.C24.0). As noted above, the intended distinction is between the 3 classes ALS, SBMA, and Healthy, which have the following intercepts and coefficients: ^ For ALS: β0 = -1.319, β1 = 0.711, β2 = -0.334, β3 = 0.718, β4 = -0.913, β5 = 0.051, β6 = 0.187, β7 = -0.203, β8 = 1.816, β9 = 0.24, β10 = 0.8, β11 = 0.675, β12 = -1.793, β13 = -0.256, β14 = -0.633, β15 = 1.506 ^ For SBMA: β0 = -1.502, β1 = 1.73, β2 = -0.568, β3 = 0.357, β4 = -0.311, β5 = - 0.763, β6 = 1.546, β7 = -0.161, β8 = -0.979, β9 = 0.842, β10 = -0.516, β11 = - 0.027, β12 = -0.706, β13 = 1.051, β14 = 0.713, β15 = -0.39 ^ For Healthy: β0 = 0, β1 = 0, β2 = 0, β3 = 0, β4 = 0, β5 = 0, β6 = 0, β7 = 0, β8 = 0, β9 = 0, β10 = 0, β11 = 0, β12 = 0, β13 = 0, β14 = 0, β15 = 0 Now, given a test sample with the feature measurements of: SM.C15.0 = 0.78, SM.C16.1 = 0.41, SM.C16.0 = 0.96, SM.C.18.1 = 0.73, SM.C.18.0 = 0.49, SM.C.20.1 = 0.46, SM.C20.0 = 0.39, SM.C21.0 = 0.76, SM.C22.1 = 0.43, SM.C22.0 = 0.36, SM.C23.1 = 0.33, SM.C23.0 = 0.25, SM.C24.2 = 0.59, SM.C24.1 = 0.67, SM.C24.0 = 1.43, the probability of this new sample belonging to each of the three classes is calculated as follows: First, we calculated the exponentiated linear predictors for each class: ^ For ALS: exp (-1.319 + 0.711(0.78) + -0.334(0.41) + 0.718(0.96) + -0.913(0.73) + 0.051(0.49) + 0.187(0.46) + -0.203(0.39) + 1.816(0.76) + 0.24(0.43) + 0.8(0.36) + 0.675(0.33) + -1.793(0.25) + -0.256(0.59) + -0.633(0.67) + 1.506(1.43)) = 9.753 ^ For SBMA: exp (-1.502 + 1.73(0.78) + -0.568(0.41) + 0.357(0.96) + -0.311(0.73) + -0.763(0.49) + 1.546(0.46) + -0.161(0.39) + -0.979(0.76) + 0.842(0.43) + - 0.516(0.36) + -0.027(0.33) + -0.706(0.25) + 1.051(0.59) + 0.713(0.67) + - 0.39(1.43)) = 0.812 ^ For Healthy: exp (0 + 0(0.78) + 0(0.41) + 0(0.96) + 0(0.73) + 0(0.49) + 0(0.46) + 0(0.39) + 0(0.76) + 0(0.43) + 0(0.36) + 0(0.33) + 0(0.25) + 0(0.59) + 0(0.67) + 0(1.43)) = 1 Next, we calculated the sum of the exponentiated linear predictors for all classes: ^ 9.753 + 0.812 + 1 = 11.565 Finally, we calculated the probability of the sample belonging to each class: ^ For ALS, : 9.753 / 11.565 = 0.84 = 84% ^ For SBMA, : 0.812 / 11.565 = 0.07 = 7% ^ For Healthy, : 1 / 11.565 = 0.09 = 9% Therefore, in this worked example, it is determined that the subject is most likely to have ALS, and a clinician may diagnose them as such, as this is the greater of the probabilities of the three classes. In addition, the predicted value of 84% gives the clinician a precise idea of how likely it is for the patient to have ALS, and not one of the other two disease conditions (SBMA or Healthy), thereby aiding the clinician in making a diagnostic decision. The same analysis (i.e. same formula, intercepts and coefficients) was then performed on a test set of 148 subjects (N = 59 Healthy; 58 ALS; 31 SBMA), and a predictive diagnosis provided for each subject according to which class (i.e. healthy, ALS, or SBMA) scored the highest value. All subjects were correctly classified into the correct disease group (sensitivity of 1 and specificity of 1). Example 5 - Predictive multinomial modelling using a selection of SM panels to predict whether a subject is healthy or whether the subject has ALS or SBMA Multinomial models were created based on a selection of different SM panels, and the corresponding data are presented in Table 2. A diagnostic accuracy of 95% was achieved with a panel of 2 SM observations (C15:0 & C21:0) – see Set E, and it was readily possible to dial-up this level of diagnostic accuracy to 96% by the addition of one further SM observation (C22:0) – see Set D (Table 2). Whilst the addition of 5 further SM observations (= 8 SM observations in total) dialled-up the accuracy level to 99% - see Set B, the addition of yet further 5 SM observations (=13 SM observations in total) provided no observable improvement - see Set A (Table 2). A diagnostic accuracy of 93% was achieved with a panel of 7 SM observations - see Set C, and it was readily possible to dial-up this level of diagnostic accuracy to 99% by the addition of one further SM observation (C21:0) – see Set C*, and / or to 100% by the addition of seven further SM observations - see Set A (Table 2). Model type and Sphingomyelins included in model ROC- Accuracy* task AUC Set A = all 15 SMs listed in Example 3 1.000 100% 5 Set A’ = Set A minus SM observations C16:0 and C18:0 1.000 99% Set B = Multinomial C15:0, C16:1, C20.0, C21:0, C22:1, C22:0, 99% regression C23:1, C23:0 to classify: 10 ALS Set C = v C16:0, C18:1, C18:0, C20:1, C24:2, C24:1, SBMA C24:0 0.980 93% v Healthy Set C’ = Set C plus SM observation C21:0 1.000 99% Set D = 15 C15:0, C21:0, C22:0 0.992 96% Set E = 0.990 95% C15:0 & C21:0 Table 2. *Accuracy = percentage of subjects correctly classified Set A = SM observations C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1,.C24.0; Set A* = panel A, excluding C16:0 & C18:0; Set B = SM observations C15:0, C16:1, C20:0, C21:0, C22:1, C22:0, C23:1 & C23:0; Set C = SM observations C16:0, C18:1, C18:0, C20:1, C24:2, C24:1 & C24:0; Set C* = panel C plus SM observation C21:0; Set D = SM observations C15:0, C21:0 & C22:0; Set E = SM observations C15:0 & C21:0 Example 6 - Predictive multinomial modelling using a selection of different SM panels to predict whether a subject is healthy or whether the subject has ALS or SBMA ANOVA p-value testing for significant differences (for ALS v SBMA v Healthy) in the mean values of individual SMs are shown below, alongside their Fold-changes (FC) for each of the three comparisons: ALS v Healthy, ALS v SBMA, and Healthy v SBMA (Table 3). Models were created using a selection of SMs based on a combination of ANOVA p-value and a threshold Fold-change. In this regard, Set A includes 15 SMs (C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1,.C24.0) having ANOVA p-value < 0.05. Set B is the subset of Set A having an absolute Fold- change greater than 2 in at least one comparison. Set C comprises C15:0 and / or C21:0, which were identified as the top significant SMs (i.e. of the SMs having ANOVA p < 0.05) for ALS v Healthy (C15:0), ALS v SBMA (C21:0), and Healthy v SBMA (C15:0, again). Table 3. SM observations demonstrating at least a 2-fold absolute (i.e. ignoring whether the change is positive or negative) expression level changes are highlighted in red (and the 3 greatest-fold changes in bold). SM observations demonstrating a p-value less than 0.05 are highlighted in green. When, for a given model, its formula was applied to a test set comprising the SM measurements of a separate and different cohort of 148 subjects (N = 59 Healthy; 58 ALS; 31 SBMA), the models had different levels of performance, here reflected by greater (better) or lower (worse) accuracy and / or ROC-AUC (Table 4). Referring to Table 4, multinomial modelling with an SM panel comprising either C15:0 (Set C’) or C21:0 (Set C’’) achieved a diagnostic accuracy of 78% and 76%, respectively. Moreover, by simply combining these two SMs to form a panel comprising C15:0 & C21:0, a diagnostic accuracy of 95% (Set C) was achieved. Model type ROC- Sphingomyelins included in model Accuracy* and task AUC Set A all 15 SMs listed in Example 3 1.000 100% (ANOVA p-value < 0.05). 15 Multinomial regression to Set B classify: C15:0, C16:0, C20.1, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0 ALS 1.000 100% (p < 0.05 plus an 20 absolute Fold-change > 2) v Set C SBMA C15:0 & C21:0 25 (p > 0.05 plus the top SM by Fold-change) 0.990 95% v Set C’ Healthy 0.833 78% C15:0 Set C’’ 0.814 76% 30 C21:0 Table 4. *Accuracy = percentage of subjects correctly classified. Table 4 confirms that the above-noted diagnostic accuracy of 95% (Set C) was yet further readily dialled-up by the inclusion of 9 additional SM observations (Set B). In this regard, Set B achieved a diagnostic accuracy of 100%, For completeness, a panel comprising all 15 of the SMs listed in Example 3 was also tested (Set A), and demonstrated a diagnostic accuracy of 100%. Example 7 - Extension to other neuromuscular disorders Samples from Spinal muscular atrophy (SMA), Parkinson’s disease (PD) and Alzheimer’s disease (AD) subjects were added to the analysis. Data are shown for levels of each individual SM (Figure 3 A-O). Different profiles are evident, specific to each disease, showing that SM panels of the present invention can distinguish between ALS, SBMA, SMA, AD, PD and healthy. Multinomial logistic models built on data either from a set of 15 SMs or from a subset of 5 SMs (C15:0, C24:2, C21:0, C22:0, C24:1) obtained perfect classification of ALS (n=60), Healthy (n=61), Parkinson’s disease (n=16), SBMA (n=60), and SMA (n=10) (Figure 4). To explore whether this strong disease specificity of EV SMs could be present in other neurological disorders, patient samples available in a Northern Irish MND Biobank (REC 21 / NI / 0010) were analysed. This initial analysis (Figure 5 A-F) indeed identified the existence of disorder specific SM profiles based on the differential expression of one or more SM observation(s), These data confirm the present invention has broad application across the genus of neurological disorders including, but not limited to: Amyotrophy Lateral Sclerosis (ALS); Charcot-Marie-Tooth (CMT); Facio Scapulohumeral Myopathy (FSHD); Myasthenia Gravis (MG); Parkinson’s disease (PD); Spinal and Bulbar Muscular Atrophy (SBMA); Spinal Muscular Atrophy type II (SMAII); Spinal Muscular Atrophy type III / IV (SMA III / IV); Inclusion Body Myositis (IBM); and Alzheimer’s disease (AD). Example 8 - A worked example using binomial statistical analysis to provide predictive diagnosis of a subject having Parkinson’s disease (PD) versus healthy A binomial logistic regression model was created to predict the probability of a subject being Healthy or Parkinson’s Disease (PD). This was based on the differential expression measurements of a subpopulation sphingomyelins isolated from EVs in a training set of 72 subjects each being either heathy or having PD (N = 64 Healthy; 8 PD). From this model was derived a mathematical formula to obtain the probability of an individual being either heathy or having PD, given as input the SM measurements in the EVs of serum from that individual. Having created a logistic regression model of this type, and taking as features the EV sphingomyelin (SM) measurements of a given serum sample from an individual patient or subject, and as prediction classes the disease conditions that one wishes to distinguish (in this example, Healthy versus PD), then by way of using the features to determine the class of a given sample, the following formula is used to obtain the probability that the individual has a given disease condition: where P(y=i|x) is the probability of the sample belonging to class i given the values of its features x1, x2, x3, up to the number of features p, and β0, β1, β2, and β3are the intercept and coefficients, respectively, up to the number of coefficients (also p, there being one coefficient per feature). This Example illustrates use of an SM panel comprising 2 different SM observations (SM.C23.1 & SM.C24.0). As noted above, the intended distinction is between the 2 classes PD and Healthy, which have the following intercepts and coefficients: ^ for PD:β0 = 64.954, β1 = -55.568, β2 = -25.924. Now, given a new sample with the feature measurements of: SM.C23.1 = 0.69 & SM.C24.0 = 1.53, the probability of the sample belonging to the class PD is calculated as follows: = 0.99999785. Therefore, in this example, it is determined that the patient is likely to have Parkinson’s Disease, and a clinician may diagnose them as such. In addition, the predicted value of 99.99% gives the clinician a precise idea of how likely it is for the patient to have PD, thereby aiding the clinician in making a diagnostic decision. Example 9 - Predictive binomial modelling using a selection of SM panels to predict whether a subject is healthy or has PD Binomial models were created based on a selection of different SM panels, and the corresponding data are presented in Table 5. Models were tested using the following SM observations C21:0 (alone), C23:1 (alone), or C24:0 (alone) with a diagnostic accuracy of 97.2%, 98.6% and 98.6% being achieved, respectively (Table 5). Moreover, by combining C21:0 & C24:0, the diagnostic accuracy was readily dialled-up to 98.6%, and by combining C21:0 & C23:1 or C23:1 & C24:0 to 100% (Table 5). Table 5 confirms a model using all 15 SMs gave perfect ROC-AUC and accuracy.

[0003] Model type and Sphingomyelins included in ROC-AUC Accuracy* task model C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, 1.000 100% Logistic regression C23.1, C23.0, C24:2, C24:1, C24.0 to classify: PD C21:0 0.996 97.2% v Healthy C23:1 0.998 98.6% C24:0 1.000 98.6% C21:0 & C24:0 1.000 98.6% C21:0 & C23:1 1.000 100% C23:1 & C24:0 1.000 100% Table 5. *Accuracy = percentage of subjects correctly classified Example 10 - Predictive binomial modelling using a selection of different SM panels to predict whether a subject is healthy or has PD Logistic linear regression models were created to predict the probability of a subject being Healthy or having Parkinson’s disease (PD) based on the measurements of certain extracellular vesicle (EV) sphingomyelins in a training set of 72 subjects each having one of these two disease conditions (N = 64 Healthy; 8 PD). From each model was derived a mathematical formula to obtain the probability of an individual having each disease condition, given as input the SM measurements in the EVs of serum from that individual. Models were created based on a selection of SM observations demonstrating a minimum p-value of p < 0.05 in combination with demonstration of the 3 greatest-fold absolute (i.e. ignoring whether the change is positive or negative) expression level changes, and SM panels then selected based on a combination of p-value and fold-change thresholds. Fold-change T-test p-value SM.C15.0 1.20 2.39E-01 SM.C16.1 1.86 8.87E-10 ** SM.C16.0 -1.64 5.21E-07 ** SM.C18.1 1.11 1.13E-01 SM.C18.0 1.07 4.17E-01 SM.C20.1 -3.56 * 1.33E-10 ** SM.C20.0 3.52 * 2.36E-16 ** SM.C21.0 -3.70 * 3.93E-09 ** SM.C22.1 -1.27 1.31E-01 SM.C22.0 -2.69 * 2.85E-07 ** SM.C23.1 -4.24 * 4.64E-10 ** SM.C23.0 2.77 * 2.06E-15 ** SM.C24.2 -1.53 3.85E-06 ** SM.C24.1 -1.30 1.74E-03 ** SM.C24.0 -3.17 * 6.03E-10 ** Table 6. SM observations demonstrating at least a 2-fold absolute (i.e. ignoring whether the change is positive or negative) expression level change are tagged * (the 3 greatest- fold expression level changes in bold). SM observations demonstrating a p-value less than 0.05 are tagged **. When, for a given model, its formula was applied to a test set comprising the SM measurements of a separate and different cohort of 72 subjects (N = 64 Healthy; 8 PD), the models had different levels of performance, here reflected by greater (better) or lower (worse) accuracy and / or ROC-AUC (Table 7).

[0004] Model type Sphingomyelins included in model ROC-AUC Accuracy* and task C20:1 alone 1.000 97.2% C21:0 alone 0.996 97.2% C23:1 alone 0.998 98.6% SM panel = C20:1, C21:0, C23:1 based on SM observations having T-test P-value < 0.05 plus 1.000 100% Top 3 SMs by Absolute Fold-Change. SM panel = C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C:24.0 based on SM 1.000 100% Logistic observations having T-test p-value < 0.05 plus regression Absolute Fold-Change greater than 2. to classify PD SM panel = C16:1, C16:0, C20:1, C20:0, v C21:0, C22:0, C23:1, C23:0, C24:2, C24:1, Healthy 1.000 100% C:24.0 based on SM observations having T- test p-value < 0.05 SM panel = C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, 1.000 100% C23.1, C23.0; C24:2, C24:1,.C24.0 Table 7. *Accuracy = percentage of subjects correctly classified As can be seen in Table 7, models using any one of the top 3 significant SMs gave near- perfect diagnostic accuracy., which could be readily dialled-up to a diagnostic accuracy of 100% by simple combination of all three SM observations (i.e. C20:1, C21:0 & C23.1).. Models comprising all SMs having p-value < 0.05 and Absolute Fold-Change greater than 2, or simply having p-value < 0.05, also gave perfect accuracy, as did a model using 15 SMs (C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0), C23.1, C23.0; C24:2, C24:1, C24.0). Example 11 - A worked example using binomial statistical analysis to provide predictive diagnosis of a subject having Parkinson’s disease (PD) versus healthy A binomial logistic regression model was created to predict the probability of a subject being Healthy or Parkinson’s Disease (PD). This was based on the differential expression measurements of a subpopulation sphingomyelins isolated from EVs in a training set of 72 subjects each being either heathy or having PD (N = 64 Healthy; 8 PD). From this model was derived a mathematical formula to obtain the probability of an individual being either heathy or having PD, given as input the SM measurements in the EVs of serum from that individual. Having created a logistic regression model of this type, and taking as features the EV sphingomyelin (SM) measurements of a given serum sample from an individual patient or subject, and as prediction classes the disease conditions that one wishes to distinguish (in this example, Healthy versus PD), then by way of using the features to determine the class of a given sample, the following formula is used to obtain the probability that the individual has a given disease condition: where P(y=i|x) is the probability of the sample belonging to class i given the values of its features x1, x2, x3, up to the number of features p, and β0, β1, β2, and β3 are the intercept and coefficients, respectively, up to the number of coefficients (also p, there being one coefficient per feature). In this example, a panel of 3 SMs (SMs C20:1, C21:0, and C23:1) are used as features, and the intended distinction is between the 2 classes PD and Healthy. The model has the following intercept and coefficients: ^ β0 = 75.495, β1 = -63.384, β2 = -30.370, β3 = -67.184 Now, given a new sample with the feature measurements of: SM C20:1 = 0.38, SM C21:0 = 0.76, and SM C23:1 =0.58, as 10-5 Therefore, in this example, it is determined that the patient is likely to have Parkinson’s Disease, and a clinician may diagnose them as such. In addition, the predicted value of 99.99% gives the clinician a precise idea of how likely it is for the patient to have PD, thereby aiding the clinician in making a diagnostic decision. Example 12 - Comparison of predictive performance of Neurofilament Light Chain (NFL) against EV sphingomyelins (SM) for ALS v Neurodegenerative Disease Controls v Healthy Neurofilament Light Chain (NFL) is being currently developed as a gold standard biomarker for use in ALS and other neurodegenerative conditions. Here, its performance as a predictor of disease status is tested against specific embodiments of the present invention. To ensure fair comparison, quantifications were performed on subject cohorts having a similar percentage breakdown of different disease conditions. In this regard, Table 8 shows the sample numbers for each disease control used. For both (i.e, NFL & SM) datasets, disease controls included several subjects having Parkinson’s disease (PD), spinal-bulbar muscular atrophy (SBMA), or spinal muscular atrophy (SMA). NFL (N as %) SM (N as %) Healthy 33 (33.7%) 60 (33.3%) ALS 47 (48.0%) 85 (47.2%) DC (Disease Controls) 18 (18.4%) 35 (19.4%) comprising (by %): PD 6 (6.1%) 12 (6.7%) SBMA 5 (5.1%) 12 (6.7%) SMA 7 (7.1%) 11 (6.1%) Total 98 (100.00%) 180 (100.00%) Table 8 Sample numbers, N, which were used for quantification of NFL or SM, broken down by disease condition. The sample numbers are also given expressed as percentages (in parentheses) of the total number for each cohort. For each task, the cohort is randomly split 50%:50% into a training set and a test set. The linear regression model is then constructed on the training set, and the model’s prediction formula is saved. Finally, the ROC-AUC and Accuracy values are obtained when the prediction formula is applied to the test set. The performance of NFL as a predictor of disease status was then assessed by way of parallel, comparative studies in which a panel of NFL biomarkers was compared against a range of different SM panel embodiments of the present invention. Said comparative studies took the form of the following three diagnostic classification challenges: o Task 1: classify as Healthy or ALS or Disease Control (PD, SBMA, and SMA as a combined group) o Task 2: classify as Healthy or ALS o Task 3: classify as Healthy or Disease (ALS and Disease Control selected from PD, SBMA and SMA as a combined group). Model type Biomarker tested ROC- Accuracy* and task (SM panel or NFL panel) AUC NFL 0.768 64.0% Multinomial regression to classify: SM (Set A) = all 15 sphingomyelins 0.933 83.5% ALS v SM (Set B) = C15:0, C16:1, C20.0, Disease Control (ALS, PD, C21:0, C22:1, C22:0, C23:1, C23:0 0.918 80.2% SBMA or SMA combined) v Set C = C16:0, C18:1, C18:0, Healthy C20:1, C24:2, C24:1, C24:0 0.840 75.8% NFL 0.832 73.2% Binomial regression to SM (Set A) = all 15 sphingomyelins 1.000 98.6% classify: ALS SM (Set B) = C15:0, C16:1, C20.0, v C21:0, C22:1, C22:0, C23:1, C23:0 1.000 98.6% Healthy SM (Set C) = C16:0, C18:1, C18:0, C20:1, C24:2, C24:1, C24:0 0.983 95.9% NFL 0.807 66.0% Binomial regression to classify: SM (Set A) = all 15 sphingomyelins 0.999 98.9% ALS and Disease Control (PD, SBMA SM (Set B) = C15:0, C16:1, C20.0, or SMA combined) C21:0, C22:1, C22:0, C23:1, C23:0 1.000 100.0% v Healthy SM (Set C) = C16:0, C18:1, C18:0, C20:1, C24:2, C24:1, C24:0 0.990 97.8% Table 9. Comparisons of ROC-AUC and Accuracy for the classification performance of various linear regression models. *Accuracy = percentage of subjects correctly classified. In each task, the SM biomarker performance was superior (higher ROC-AUC and higher Accuracy) to the NFL biomarker in predicting disease status, thereby confirming the EV sub-population of SMs according to the present invention as the stand-out biomarker of choice. Example 13 - Predictive modelling using a selection of different SM panels to provide predictive diagnosis across combined neurological disorders versus healthy One powerful advantage of the present invention is the flexibility and ease with which a unique SM profile representative of a different neurological disorder cohort may be selected and / or combined according to the diagnostic question being posed. A model is then created (e.g. by binomial or multinomial linear regression analysis) and a mathematical formula derived to obtain the probability that an individual belongs to each of the neurological disorder cohorts, and thus to which cohort the individual has closest nexus. This is nicely illustrated by SM panels A-C (Example 12), which provide the basis for predictively diagnosing a subject (by binomial or multinomial modelling) as being healthy or as having a neurological disorder identified either as a specific diagnosis of ALS, PD, SBMA or SMA, or as a generic diagnosis to any one of ALS, PD, SBMA or SMA. Simple extension of the above diagnostic question to include AD provides the following SM profile output (binomial modelling) for any neurological disorder (ALS, PD, SBMA, SMA or AD) versus healthy: SMs demonstrating a p-value < 0.05: C15:0, C16:1, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0. SMs demonstrating aa Absolute Fold-change > 2.0 presented in descending order (highest Fold-change first), any of which is an SM observation proposed for panel consideration: C15:0, C20:1, C21:0, C22:1, C22:0, C16:0, C23:1, C24:0. Preferably any one, two or three selected from C15:0, C20:1, C21:0; more preferably any four, five or six selected from C15:0, C20:1, C21:0, C22:1, optionally C22:0, C16:0. Example 14 - Predictive modelling using a selection of different SM panels to provide predictive diagnosis between different subtypes of the same neurological disorder (ALS) The present invention has the power to distinguish between the three sub-groups of ALS. Addressing the first diagnostic question of ALS (non-PMA / PLS) v PLS provides the following SM profile output (binomial modelling): SMs demonstrating an Absolute Fold-change > 1.90 presented in descending order (highest Fold-change first), any or which is an SM observation proposed for panel consideration: C23.1, C22:1, C24:1, C21:0, C16:0 Preferably any one, two or three SM observation selected from: C23.1, C22:1, optionally C24:1; more preferably any four or five selected from: C23.1, C22:1, C24:1, C21:0, optionally C16:0. Addressing the second diagnostic question of ALS (non-PMA / PLS) v PMA provides the following SM profile output (binomial modelling): SMs demonstrating a p-value < 0.05: C16:0, C22:0, C23:0. SMs demonstrating an Absolute Fold-change > 1.20 presented in descending order (highest Fold-change first), any or which is an SM observation proposed for panel consideration: C23.0, C23:1, C18:0, C22:0, C16:0. Preferably any one, two or three SM observation selected from: C23.0, C23:1, optionally C18:0; more preferably any four or five selected from: C23.0, C23:1, C18:0, C22:0 optionally C16:0. Addressing the third diagnostic question of PLS v PMA provides the following SM profile output (binomial modelling): SMs demonstrating an Absolute Fold-change > 2.0 presented in descending order (highest Fold-change first), any or which is an SM observation proposed for panel consideration: C23:1, C24:1, C23:0, C18:0, C20:1. Preferably any one, two or three SM observation selected from: C23.1, C24:1, optionally C23:0; more preferably any four or five selected from: C23.1, C24:1, C23:0, C18:0 optionally C20:1. SM profile comparisons of familial ALS versus sporadic ALS, and of ALS (positive for a “C9” or “F72” mutation) versus ALS (negative for a “C9” or “F72” mutation) yield no statistically significant p-values (p < 0.05). This was expected and confirms that both genetic and familial cases are not behaving differently when assessed in accordance with the present invention. Example 15 - Tracking of disease severity (SMA II versus SMA type III / IV) There are differences in the pattern of SM levels between SMA type II and SMA type III / IV (Figure 6 A - G), suggesting that SM profiles may reflect the severity of SMA disease, and could be used to non-invasively monitor therapeutic efficacy – a pressing need in ongoing trials of Nusinersen. This suggests that the SM panel could be useful for disease tracking, not only for SMA, but for other conditions as well. Addressing the diagnostic question of SMA type II versus SMA type III / IV provides the following SM profile output (binomial modelling): SMs demonstrating a p-value < 0.05: C15:0, C16:0, C24:0. SMs demonstrating an Absolute Fold-change > 1.20 presented in descending order (highest Fold-change first), any of which is an SM observation proposed for panel consideration: C15:0, C16:0, C18:1, C24:1, C18:0, C24:2, C24:0. Preferably any one, two or three SM observations selected from: C15.0, C16:0, optionally C18:1, C24:0; more preferably any four or five selected from: C15.0, C16:0, C18:1, C24:1 optionally C18:0, C24:2. C24:0. Example 16 - Demonstration that the EV sphingomyelin panel responds to therapeutic treatment and can be used to track therapeutic efficacy Ongoing clinical trials (such as of the Nusinersen antisense oligonucleotide) are hampered by the lack of accurate biomarkers to track Spinal Muscular Atrophy (SMA) treatment efficacy. Effective non-invasive biomarkers could be used to monitor disease progress, making it possible to assess clinical benefit or to optimize patient selection / stratification. In SMA patients, we show that the extracellular vesicles present different SM profiles compared to healthy subjects and other neurological conditions (Figures 1, 3, and 5). In addition, considering two types of SMA, types II and III / IV, type II being a more severe form than type III / IV, there are differences in the pattern of SM levels compared to healthy (Figure 6): of several SMs that are increased compared to healthy, their extent of increase is greater for type II than for type III / IV. This indicates that SM profiles reflect the severity of SMA disease and could be used to non-invasively monitor therapeutic efficacy. To confirm that EV SM profiles correlate with therapeutic efficacy in SMA, we extracted EVs from a SMA murine model treated with an adeno-associated virus serotype 9 (AAV9)- SMN gene therapy. The SM profile were analysed in serum collected before and after treatment, and from age-matched healthy (wild-type) controls. The following six SMs were found to be at significantly increased levels in untreated SMA compared to healthy, and to show rescue towards healthy levels following treatment of the SMA murine model with the AAV9-SMN therapy: SMs C15:0, C18:1, C18:0, C20:1, C20:0, and C23:0 (Figure 10). These findings in mice suggest that the set of 14 SMs found to be at significantly different levels in the EVs of human SMA patients compared to healthy controls could also be used to track response to therapy: SMs C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0. These findings indicate that the EV SM panel could be used to monitor therapeutic efficacy in SMA and in other neurological / neuromuscular conditions. This would be applicable both to human clinical trials and to animal testing. It follows that, as well as tracking disease progress or rescue thereof, the EV SM panel could also be used as a companion diagnostic, such as to pre-select the patients thought most likely to respond to a given therapy (e.g. those patients having the more elevated levels of SMs). Example 17 - No effect of age or sex, stability over long-term storage, and consistency across collection sites SM levels were observed to be independent of age (Figure 7A) or sex (Figure 7B). SM levels are stable over long-term storage (Figure 8A) and consistent between sample collection sites (Figures 8B-F). Example 18 - A “decision Tree” (or “Selection Algorithm”) approach to the present invention The following example is to be read in conjunction with Figure 11, In this example, the EV SM levels were collected for each subject of a cohort of 113 ALS, 65 SBMA, and 128 Healthy. A “decision Tree” (or “Selection Algorithm”) was then created which progressively sub-divides the subjects based on their SM levels: each branch point (or 'threshold split') in the decision tree represents a new sub-division based on the level of a selected SM. The proportion of Healthy individuals for each threshold split is represented by a green block, ALS by red, and SBMA by blue. The decision tree can be used as a classification tool for new patients. Probabilities are indicated for whether a subject is likely to have the disorder given that they have SM levels that follow a given combination of thresholds. For example, for a sample from a new patient, a prediction could be made as to their disease status as follows: after measuring the levels of EV SMs in the new patient, one would ask if the new patient has a level of C15:0 greater than 0.32? A finding that the patient does have C15:0 > 0.32 would place that patient on the right-hand branch of the tree. It could then be asked, for the next sub-division, does the patient have C21:0 > 0.43? If the new patient has C21:0 > 0.43, this places them in the right-most sub-division. As indicated in the figure, subjects falling within this sub-division have a probability of having ALS of 0.9942 (99.42%). The key SM observations are selected based on their ability to classify the subjects. Therefore, a clinician could be confident to diagnose the new patient as having ALS, and that the patient neither has SBMA, nor is healthy. Example 19 - Predictive modelling with a selected SM panel to provide predictive diagnosis between Fast progressive ALS and Slow progressive ALS ALSFRS score is a measure of clinical severity in ALS disease. The rate of change of the ALSFRS (the dALSFRS score) is a measure of the rate at which ALS disease has progressed in a patient over time. It is calculated by assessing the ALSFRS are two time- points, calculating the amount of change, and dividing this amount by the time duration between the time-points. In this study, the dALSFRS score for a given patient was calculated at the time of blood sampling, based on that patient’s change in ALSFRS for the preceding months (usually 3 to 6 months). Each subject was then defined as a Fast progressor if they had dALSFRS score >1, and as a slow progressor if they had dALSFRS <0.5. The p-values and the Fold-changes are given (Table 10). Sphingomyelin Fold-change T-test p-value SM.C15.0 -1.07 1.77E-01 SM.C16.1 -1.23 3.17E-03 * SM.C16.0 -1.15 6.76E-02 SM.C18.1 -1.57 1.37E-07 * SM.C18.0 1.05 3.69E-01 SM.C20.1 -1.08 2.02E-01 SM.C20.0 -1.27 1.74E-03 * SM.C21.0 -1.33 3.42E-04 * SM.C22.1 1.15 3.16E-02 * SM.C22.0 -1.40 2.91E-06 * SM.C23.1 1.05 8.19E-01 SM.C23.0 1.87 8.16E-03 * SM.C24.2 1.41 3.35E-04 * SM.C24.1 -1.37 2.77E-08 * SM.C24.0 1.19 7.31E-04 * Table 10. A p-value of less than 0.05 is highlighted by way of * Logistic linear regression models were created to predict the probability of a subject being a Fast or Slow progressor based on the measurements of certain extracellular vesicle (EV) sphingomyelins in a training set of 45 subjects each having one of these two disease conditions (N = 16 Fast; 29 Slow). From each model was derived a mathematical formula to obtain the probability of an individual having each disease condition, given as input the SM measurements in the EVs of serum from that individual. When, for a given model, its formula was applied to a test set comprising the SM measurements of a separate and different cohort of 46 subjects (N = 17 Fast; 29 Slow), the models had different levels of performance, here reflected by greater (better) or lower (worse) accuracy and / or ROC-AUC (Table 11). As can be seen in Table 11, a model comprising all SMs having P-value < 0.05 gave accuracy of 83%. A model using all 15 SMs gave accuracy of 89%. This proves that the SM profile has a strong capacity to distinguish between patients having different rates of disease progression. Therefore, these results suggest that the SM profile may be useful in a predictive sense with respect to progression: i.e. that, by measuring the SM levels of a patient’s serum vesicles at a time of sampling, it will be possible to predict the rate of disease progression from that timepoint forwards. Model type ROC- Sphingomyelins included in model Accuracy* and task AUC Logistic regression SM having T-test P-value < 0.05 C16:1, C18:1, C20:0, C21:0, C22:1, 0.789 83% Fast C22:0, C23:0, C24:2, C24:1, C:24.0 v All 15 sphingomyelins 0.888 89% Slow *Accuracy = Percentage of subjects correctly classified Table 11.

Claims

CLAIMS 1. A method for diagnosing a first neurological disorder in a subject, said method comprising a) providing a patient primary sphingomyelin (SM) profile for said subject, i. wherein said patient primary SM profile is obtainable by mass spectrometry analysis of a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject, and ii. wherein said patient primary SM profile comprises one or more primary SM observation(s), each of which is a different SM observation; b) providing a reference primary SM profile comprising one or more primary SM observation(s): i. wherein said reference primary SM profile is obtainable by comparative mass spectrometry analysis of body fluid samples comprising 30- 300nm EVs isolated from two different reference cohorts selected from a healthy subject cohort and a cohort of subjects having the first neurological disorder; ii. wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD) value for said one or more primary SM observation(s); and (b) between each reference cohort, identification of one or more key SM observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and iii. wherein said reference primary SM profile comprising one or more of said key SM observation(s) is provided in a form that is suitable for comparative analysis with the patient primary SM profile; c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile such that: i. when the healthy subject cohort is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said correspondingone of more key SM observation(s) of the reference primary SM profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has the first neurological disorder; and / or ii. when the cohort of subjects having the first neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) of said corresponding one of more key SM observation(s) of the reference SM profile, said “match” being indicative the subject has the first neurological disorder and / or wherein a “non-match” is indicative the subject is healthy.

2. The method according to claim 1, said method further comprising: a. providing a patient secondary SM profile for said subject: i. wherein said patient secondary SM profile is obtainable by mass spectrometry analysis of a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject, and ii. wherein said secondary patient SM profile comprises one or more secondary SM observation(s), each of which is a different SM observation; and b. providing a reference secondary SM profile comprising one or more secondary SM observation(s): i. wherein the reference secondary SM profile is obtainable by comparative mass spectrometry analysis of body fluid samples comprising 30-300nm EVs isolated from two different reference cohorts selected from a cohort of subjects having a first neurological disorder and a cohort of subjects having a second neurological disorder, wherein the first and second neurological disorders are different; ii. wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD) value for said one or more secondary SM observation(s); and (b) between each reference cohort, identification of one or more keySM observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and iii. wherein said reference secondary SM profile comprising one or more of said key secondary SM observation(s) is provided in a form that is suitable for comparative analysis with the patient secondary SM profile; c. comparing one or more of said key SM observation(s) present in the reference secondary SM profile with the corresponding SM observation(s) present in the patient secondary SM profile such that: i. when the cohort of subjects having the first neurological disorder is selected as the reference secondary SM profile and the mean-fold expression value(s) for said one or more key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient secondary SM profile, a ”match” is defined when the value(s) for said one or more patient key secondary SM observation(s) falls within 2 SD of the mean expression value(s) of said corresponding one or more key SM observation(s) of the reference secondary SM profile, said “match” being indicative the subject has the first neurological disorder and / or a “non-match” being indicative the subject has the second neurological disorder; and / or ii. when the cohort of subjects having the second neurological disorder is selected as the reference secondary SM profile and the mean-fold expression value(s) for said one or more key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient secondary SM profile, a ”match” is defined when the value(s) for said one or more patient key secondary SM observation(s) falls within 2 SD of the mean expression value(s) of said corresponding one or more key SM observation(s) of the reference secondary SM profile, said “match” being indicative the subject has the second neurological disorder and / or a “non-match” being indicative the subject has the first neurological disorder.

3. A method for diagnosing a first or second neurological disorder in a subject, said method comprising: a. providing a patient primary sphingomyelin (SM) profile for said subject,i. wherein said patient primary SM profile is obtainable by mass spectrometry analysis of a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject, and ii. wherein said patient primary SM profile comprises one or more primary SM observation(s), each of which is a different SM observation; b. providing a reference primary SM profile comprising one or more primary SM observations: i. wherein said reference primary SM profile is obtainable by comparative mass spectrometry analysis of body fluid samples comprising 30- 300nm EVs isolated from two different reference cohorts selected from a cohort of subjects having the first neurological disorder and a cohort of subjects having the second neurological disorder; ii. wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD) value for said one or more primary SM observation(s); and (b) between each reference cohort, identification of one or more key SM observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; iii. wherein said reference primary SM profile comprising one or more of said key SM observation(s) is provided in a form that is suitable for comparative analysis with the patient primary SM profile; c. comparing said one or more key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile such that: i. when the cohort of subjects having the first neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, said “match” being indicative the subject has the first neurological disorder and / or wherein a “non-match” being indicative the subject has the second neurological disorder; and / orii. when the cohort of subjects having the second neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, said “match” being indicative the subject has the second neurological disorder and / or wherein a “non-match” being indicative the subject has the first neurological disorder..

4. The method according to claim 3, said method further comprising: a) providing a patient secondary SM profile for said subject, i. wherein said patient secondary SM profile is obtainable by mass spectrometry analysis of a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject, and ii. wherein said patient secondary SM profile comprises one or more secondary SM observation(s), each of which is a different SM observation; b) providing a reference secondary SM profile for a subject having said first neurological disorder and / or for a subject having a third neurological disorder (and that is different from the first and second disorders): i. wherein said reference secondary SM profile is obtainable by mass spectrometry analysis of body fluid samples comprising 30-300nm EVs isolated from two different reference cohorts selected from a cohort of subjects having the first neurological disorder and a cohort of subjects having said third neurological disorder; ii. wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD) value for said one or more secondary SM observation(s); and (b) between each reference cohort, identification of one or more key SM observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and iii. wherein said reference secondary SM profile comprising one or more of said key SM observation(s) is provided in a form that is suitable for comparative analysis with the patient second SM profile;c) comparing the patient secondary SM profile with the reference secondary SM profile such that: i. when the cohort of subjects having the first neurological disorder is selected as the reference secondary SM profile and the mean-fold expression value(s) for said one or more key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient secondary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference secondary SM profile, said “match” being indicative the subject has the first neurological disorder and / or a “non-match” being indicative the subject has the third neurological disorder; and / or ii. when the cohort of subjects having the third neurological disorder is selected as the reference secondary SM profile and the mean-fold expression value(s) for said one or more key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient secondary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference secondary SM profile, said “match” being indicative the subject has the third neurological disorder and / or a “non-match” being indicative the subject has the first neurological disorder;.

5. The method according to any preceding claim, wherein diagnosis is performed by binomial logistic regression modelling and by reference to the following formula:where P (y = i | x) is the probability of the sample belonging to class i given the values of its features x1, x2, x3, up to the number of features p, and β0, β1, β2, and β3are the intercept and coefficients, respectively, up to the number of coefficients (also p, here being one coefficient per feature). a) wherein class i = a subject having a first neurological disorder or a healthy subject; orb) wherein class i = a subject having a first neurological disorder or a subject having a second neurological disorder; c) and each of x1, x2, … xp = the absolute expression level value measured in the test sample for each SM observation (n = 1 to n = p) that has been selected for inclusion in the diagnostic method.

6. The method according to any preceding claim, wherein diagnosis is performed by multinomial regression analysis and by reference to the following formula:where P (y = i | x) is the probability of the sample belonging to class i given the values of its features x1, x2, x3, up to the number of features p, and β0i, β1i, β2i, and β3i are the intercept and coefficients, respectively, for class i, up to the number of coefficients (also p, here being one coefficient per feature). The denominator of the equation is the sum of the exponentiated linear predictors for all the classes from j to the total number of classes k. a) wherein class i = a subject having a first neurological disorder, a healthy subject or a subject having a second neurological disorder, optionally including a third and / or further neurological disorder; or b) wherein class i = a subject having a first neurological disorder, a subject having a second neurological disorder or a subject having a first neurological disorder, optionally including a fourth and / or further neurological disorder; c) and each of x1, x2, … xp = the absolute expression level value measured in the test sample for each SM observation (n = 1 to n = p) that has been selected for inclusion in the diagnostic method.

7. The method of any one of claims 1-6, wherein the SM comprises: a) a fatty acid base selected from d16:1, d17:1; d18:1; d19:1; d20:1; d43:1; d34:1; d32:0; d38:3; d33:0; d38:2; d33:1; d32:1; d44:1; d42:1; d41:1; d35:1; d36:1; d44:2; d34:2; and d36:3, preferably a fatty acid base selected from d16:1, d17:1, d18:1, d19:1, and d20:1; and wherein b) the SM is selected from SM14:0, SM14:1, SM15:0, SM16:0, SM16:1, SM17:0, SM18:0, SM18:1, SM19:0, SM20:0, SM20:1, SM20:2, SM21:0, SM22:0, SM22:1, SM23:0, SM23:1, SM24:0, SM24:1, SM24:2, SM25:0, SM25:1, SM26:0 and SM26:1, preferably wherein the SM is selected fromSM14:0, SM15:0, SM16:0, SM16:1, SM17:0, SM18:0, SM18:1, SM19:0, SM20:0, SM20:1, SM21:0, SM22:0, SM22:1, SM23:0, SM23:1, SM24:0, SM24:1, SM24:2, SM25:0, SM25:1, SM26:0 and SM26:

1.

8. The method of any one of claims 1-2 or 5-7, wherein the neurological disorder is Alzheimer’s disease (AD) and the patient and reference SM profile(s) share in common one or more key SM observation selected from: C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C23:1, C24:2, C24:1, C24:0; preferably from: C15:0, C18:1, C21:0, C24:2, C24:1; and more preferably from C15:0, C24.2, optionally C24.

1.

9. The method of any one of claim 8, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from:C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C23:1, C24:2, C24:1, C24:0; preferably from: C15:0, C18:1, C24:2, C24:1, optionally C21:

0.

10. The method of any one of claims 1-2 or 5-7, wherein the neurological disorder is amyotrophic lateral sclerosis (ALS) and the “patient” and “reference” SM profile(s) share in common one or more key SM observation(s) selected from: C15:0, C16:1, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22.0, C23:1, C23:0, C24:2, C24:1, and C24:0; preferably from: C15:0, C16:0, C20:1, C21:0, C22:1, C22.0, C23:0, C24:0; and more preferably from C15, C21:0, optionally C23:0 11. The method of claim 10, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C23:1, C24:2, C24:1, C24:0; preferably from: C15:0, C16:0, C20:1, C21:0, C22:1, C22.0, C23:0, C24:0; and more preferably C15:0, C21:0, C23:0, C22:0, optionally C22:

1.

12. The method of any one of claims 1-2 or 5-7, wherein the neurological disorder is myasthenia gravis (MG) and the “patient” and “reference” SM profile(s) share in common one or more key SM observation(s) selected from: C15:0, C16:1, C16:0, C18:1, C20:1, C21:0, C22:1, C23:1, C23:0, C24:2, C24:1, C24:0; preferably from: C15:0, C16:1, C16:0, C18:1, C20:1, C21:0, C23:1, C23:0, C24:2, C24:0;and more preferably from: C21:0, C23:0, optionally C23.

1.

13. The method of claim 12, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C15:0, C16:1, C16:0, C18:1, C20:1, C21:0, C22:1, C23:1, C23:0, C24:2, C24:1, C24:0; preferably from: C15:0, C16:1, C16:0, C18:1, C20:1, C21:0, C23:1, C23:0, C24:2, C24:0; and more preferably from C21:0, C23:0, C23.1, C15:0, optionally C20:

1.

14. The method of any one of claims 1-2 or 5-7, wherein the neurological disorder is Parkinson’s disease (PD) and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C16:1, C16:0, C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0, or C16:0, C20:0, C20:1, C21:0, C22:0, C23:0, C23:1, C24:0, C24:2; preferably from: C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:0; and more preferably from: C23:1, C21:0, optionally C20:

1.

15. The method of claim 14, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C16:1, C16:0 , C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0; preferably from: C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:0; and more preferably from C23:1, C21:0, C20:1, C20:0, optionally C24:

0.

16. The method of any one of claims 1-2 or 5-7, wherein the neurological disorder is progressive muscular atrophy (PMA) and the “patient” and “reference” SM profile(s) share in common one or more key SM observation(s) selected from: C15:0, C16:0, C20:0, C21:0, C22:0, C23:0, C24:0; and preferably from: C21:0, C23:0, optionally C15:

0.

17. The method of claim 16, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C15:0, C16:0, C20:0, C21:0, C22:0, C23:0, C24:0; and preferably from: C21:0, C23:0, C15:0, C22:0, optionally C16:0.

18. The method of any one of claims 1-2 or 5-7, wherein the neurological disorder is spinal and bulbar muscular atrophy (SBMA) and the “patient” and “reference” SM profile(s) share in common one or more key SM observation(s) selected from: C15:0, C16:1, C16:0, C18:1, C18:0, C20:1, C20:0, C22:1, C23:0, C24.2, C24.1, C24:0, or C15:0, C16:0, C16:1, C18:0, C18:1, C20:0, C20:1, C22:0, C22:1, C23:1, C24:0, C24:1, C24:2; or C15:0, C16:0, C18:0, C20:1, C20:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1; preferably from: C15:0, C22:1, C20:1, C24.2, C16:0, C23.1, C24.1; and more preferably from: C15:0, C22:1, optionally C20:

1.

19. The method of claim 18, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C15:0, C16:1, C16:0, C18:1, C18:0, C20:1, C20:0, C22:1, C23:0, C24.2, C24.1, C24:0, or C15:0, C16:0, C16:1, C18:0, C18:1, C20:0, C20:1, C22:0, C22:1, C23:1, C24:0, C24:1, C24:2; or C15:0, C16:0, C18:0, C20:1, C20:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1; preferably from: C15:0, C22:1, C20:1, C24.2, C16:0, C23.1, C24.1; and more preferably: C15:0, C22:1, C20:1, C24.2, optionally C16:0 or C23.

1.

20. The method of any one of claims 1-2 or 5-7, wherein the neurological disorder is spinal muscular atrophy (SMA) and the “patient” and “reference” SM profile(s) share in common one or more key SM observation(s) selected from: C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24.2, C24.1, C24:0, or C15:0, C16:0, C18:0, C18:1, C20:1, C22:0, C22:1, C24:0, C24:2; preferably from: C15:0, C16:0, C20:1, C22:1; and more preferably from: C15:0, C22:1, optionally C20:

1.

21. The method of claim 20, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C15:0, C16:0, C18:1, C18:0, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24.2, C24.1, C24:0, or C15:0, C16:0, C18:0, C18:1, C20:1, C22:0, C22:1, C24:0, C24:2; preferably from: C15:0, C16:0, C20:1, C22:1; optionally C18:1.

22. The method according to any one of claims 3-7, for distinguishing a subject having Fast progression ALS from a subject having Slow progression ALS, wherein the reference SM profiles compared are Fast versus Slow progression ALS, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C16:1, C18:1, C20:0, C21:0, C22:1, C22:0, C23:0, C24.2, C24.1, C24:0; preferably from C16:1, C18:1, C20:0, C21:0, C22:0, C23:0, C24.2, C24.1; more preferably from C18:1, C23:0, optionally C24:

2.

23. The method according to claim 22, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C16:1, C18:1, C20:0, C21:0, C22:1, C22:0, C23:0, C24.2, C24.1, C24:0; preferably from C16:1, C18:1, C20:0, C21:0, C22:0, C23:0, C24.2, C24.1; more preferably from C18:1, C22:0, C23:0, C24:2, optionally C24:

1.

24. The method according to any one of claims 3-7, for distinguishing a subject having Severe ALS from a subject having Mild ALS, wherein the reference SM profiles compared are Severe versus Mild ALS, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C24.2, C24.0, C23:0, C20:0, C18:1; preferably C24:2, C24:0, optionally C23:

0.

25. The method according to claim 24, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C24.2, C24.0, C23:0, C20:0, optionally C18:

1.

26. The method according to any one of claims 3-7, for distinguishing a subject having AD from a subject having ALS, wherein the reference SM profiles compared are AD versus ALS, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C24.2, C24.1, C18:1, C22:0, C22:1, C23:0, C16:0, C20:1, C21:0, C16:1, C18:0, C20:0, C24:0; or C16:0, C18:1, C20:1, C21:0, C22:0, C22:1, C23:0, C24:0, C24:1, C24:2; preferably from: C24.2, C24.1, C18:1, C22:0, C22:1, C23:0, C16:0, C20:1; more preferably from: C24.2, C24.1, optionally C18:1.

27. The method according to claim 26, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C24.2, C24.1, C18:1, C22:0, C22:1, C23:0, C16:0, C20:1, C21:0, C16:1, C18:0, C20:0, C24:0, or C16:0, C18:1, C20:1, C21:0, C22:0, C22:1, C23:0, C24:0, C24:1, and C24:2; preferably from: C24.2, C24.1, C18:1, C22:0, C22:1, C23:0, C16:0, C20:1; more preferably from: C24.2, C24.1, C18:1, C20:0, optionally C22:

1.

28. The method according to any one of claims 3-7, for distinguishing a subject having AD from a subject having MG, wherein the reference SM profiles compared are AD versus MG, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C24:1, C23:1, C18:1, C23:0, C21:0, C20:1, C24:0, C16:0, C16:1, C18:0; preferably from: C16:0, C24:1, C23:1, C18:1, C23:0, C21:0, C20:1, C24:0; more preferably from: C24:1, C23:1, optionally C18:

1.

29. The method according to claim 28, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C24:1, C23:1, C18:1, C23:0, C21:0, C20:1, C24:0, C16:0, C16:1, C18:0; preferably from: C16:0, C24:1, C23:1, C18:1, C23:0, C21:0, C20:1, C24:0; more preferably from: C24:1, C23:1, C18:1, C23:0, optionally C21:

0.

30. The method according to any one of claims 3-7, for distinguishing a subject having AD from a subject having PD, wherein the reference SM profiles compared are AD versus PD, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C24:2, C15:0, C24:1, C20:0, C23:1, C22:0, C20:1, C24:0, C23:0, C18:0, C16:1, C21:0, C16:0; or C15:0, C18:1, C20:1, C20:0, C22:0, C23:1, C23:0, C24:0, C24:1, C24:

2. preferably from: C24:2, C15:0, C24:1, C20:0, C23:1, C22:0; more preferably from: C24:2, C15:0, optionally C24:

1.

31. The method according to claim 30, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C24:2, C15:0, C24:1, C20:0, C23:1, C22:0, C20:1, C24:0, C23:0; or C15:0, C18:0, C18:1, C20:1, C20:0, C22:0, C23:1, C24:0, C24:1, C24:

2. preferably from: C24:2, C15:0, C24:1, C20:0, optionally C23.

32. The method according to any one of claims 3-7, for distinguishing a subject having AD from a subject having PMA, wherein the reference SM profiles compared are AD versus PMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C24:2, C18:1, C23:0, C20:0, C21:0, C22:0, C18:0, C16:0, C24:0, C20:1; preferably from: C24:2, C18:1, C23:0, C20:0, C21:0, C22:0; more preferably from: C24:2, C18:1, optionally C23:

0.

33. The method according to claim 32, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C24:2, C18:1, C23:0, C20:0, C21:0, C22:0, C18:0, C16:0, C24:0, C20:1; preferably from: C24:2, C18:1, C23:0, C20:0, C21:0, C22:0; more preferably from: C24:2, C18:1, C23:0, C20:0, optionally C21:

0.

34. The method according to any one of claims 3-7, for distinguishing a subject having AD from a subject having SBMA wherein the reference SM profiles compared are AD versus SBMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C24.2, C24:1, C18:1, C22:1, C20:1, C21:0, C16:0, C23:1, C16:1, C18:0, C24:0; or C16:0, C18:0, C18:1, C20:1, C21:0, C22:1, C24:0, C24:1, C24:2; preferably from: C24.2, C24:1, C18:1, C22:1, C20:1, C21:0, C16:0; more preferably from: C24.2, C24:1, optionally C18:

1.

35. The method according to claim 34, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C24.2, C24:1, C18:1, C22:1, C20:1, C21:0, C16:0, C23:1, C16:1, C18:0, C24:0; or C16:0, C18:0, C18:1, C20:1, C21:0, C22:1, C24:0, C24:1, C24:

2. preferably from: C24.2, C24:1, C18:1, C22:1, C20:1, C21:0, more preferably from: C24.2, C24:1, C18:1, C22:1, optionally C20:

1.

36. The method according to any one of claims 3-7, for distinguishing a subject having AD from a subject having SMA wherein the reference SM profiles compared are AD versus SMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C24:1, C24:2, C18:1, C22:1, C16:0, C20:1, C22:0, C21:0, C16:1, C24:0, or C15:0, C16:0, C18:1, C20:1, C22:1, C24:2, C24:0;preferably from: C24:1, C24:2, C18:1, C22:1, C16:0, C20:1; more preferably from: C24:1, C24:2, optionally C18:1..

37. The method according to claim 36, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C24:1, C24:2, C18:1, C22:1, C20:1, C16:0, C22:0, C21:0, C16:1; or C15:0, C16:0, C18:1, C20:1, C22:1, C24:2, C24:0; preferably from: C24:1, C24:2, C18:1, C22:1, optionally.C16:

0.

38. The method according to any one of claims 3-7, for distinguishing a subject having ALS from a subject having MG wherein the reference SM profiles compared are ALS versus MG, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C23:1, C22:1, C24:2, C16:1, C22:0, C23:0, C21:0, C18:1, C18:0, C24:0, C20:0; preferably from: C23:1, C22:1, C24:2, C16:1, C22:0, C23:0, C18:1, C21:0; more preferably from: C23:1, C22:1, optionally C24:

2.

39. The method according to claim 38, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C23:1, C22:1, C24:2, C16:1, C22:0, C23:0, C21:0, C18:1, C18:0, C24:0, C20:0; preferably from: C23:1, C22:1, C24:2, C16:1, C22:0, C23:0, C21:0, C18:1; more preferably from: C23:1, C22:1, C24:2, C16:1, optionally C22:

0.

40. The method according to any one of claims 3-7, for distinguishing a subject having ALS from a subject having PD wherein the reference SM profiles compared are ALS versus PD, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15:0, C20:0, C16:1, C23:1, C22:1, C21:0, C16:0, C18:1, C24:0, C23:0, C18:0, C20:1, C22:0; or C15:0, C16:0, C16:1, C18:0, C20:0, C20:1, C22:1, C23:1, C24:0; preferably from: C15:0, C20:0, C16:1, C23:1, C22:1; more preferably from: C15:0, C20:0, optionally C16:

1.

41. The method according to claim 40, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C15:0, C20:0, C16:1, C23:1, C22:1, C21:0, C16:0, C18:1, C24:0, C23:0, C18:0, C20:1, C22:0; orC15:0, C16:0, C16:1, C18:0, C20:0, C20:1, C22:1, C23:1, C24:0; preferably from: C15:0, C20:0, C16:1, C23:1, C22:1, C21:0, C16:0, C18:1; more preferably from: C15:0, C20:0, C16:1, C23:1, optionally C22:

1.

42. The method according to any one of claims 3-7, for distinguishing a subject having ALS from a subject having PMA wherein the reference SM profiles compared are ALS versus PMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C23:0, C16:0, C22:0, C20:0, C16:1, C22:1; preferably from: C23:0, C16:0, optionally C22:

0.

43. The method according to claim 42, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C23:0, C16:0, C22:0, C20:0, optionally C16:

1.

44. The method according to any one of claims 3-7, for distinguishing a subject having ALS from a subject having SBMA wherein the reference SM profiles compared are ALS versus SBMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C21:0, C22:0, C23:0, C20:1, C22:1, C24:2, C24:1, C15:0, C24:0, C23:1, C20:0, C16:0, or C15:0, C18:0, C18:1, C20:1, C21:0, C22:0, C22:1, C23:0, C24:0, C24:1, C24:2; preferably from: C21:0, C22:0, C23:0, C20:1, C22:1, C24:2; or preferably from: C15:0, C18:0, C20:1, C20:0, C21:0, C22:1, C23:1, C24:2; more preferably from: C21:0, C22:0, optionally C23:

0.

45. The method according to claim 40, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C21:0, C22:0, C23:0, C20:1, C22:1, C24:2, C24:1, C15:0, C24:0, C23:1, C20:0, C16:0, or C15:0, C18:0, C18:1, C20:1, C21:0, C22:0, C22:1, C23:0, C24:0, C24:1, C24:2; preferably from: C21:0, C22:0, C23:0, C20:1, C22:1, C24:2, C24:1, C16:0; or preferably from: C15:0, C18:0, C20:1, C20:0, C21:0, C22:1, C23:1, C24:2; more preferably from: C21:0, C22:0, C23:0, C20:1, optionally C22:

1.

46. The method according to any one of claims 3-7, for distinguishing a subject having ALS from a subject having SMA wherein the reference SM profiles comparedare ALS versus SMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C21:0, C23:0, C15:0, C22:0, C24:0, C16:1, C16:0, C24.1; or C15:0, C18:0, C18:1, C21:0, C22:0, C23:0, C24:0, C24:1; preferably from: C21:0, C23:0, C15:0, C22:0, C24:0, C16:1; more preferably from: C21:0, C23:0, optionally C15:

0.

47. The method according to claim 46, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C21:0, C23:0, C15:0, C22:0, C24:0, C16:1, C16:0, C24.1; or C15:0, C18:0, C18:1, C21:0, C22:0, C23:0, C24:0, C24:1; preferably from: C21:0, C23:0, C15:0, C22:0, C24:0, C16:1; more preferably from: C21:0, C23:0, C15:0, C22:0, optionally C24:

0.

48. The method according to any one of claims 3-7, for distinguishing a subject having MG from a subject having PD wherein the reference SM profiles compared are MG versus PD, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C23:1, C15:0, C24:2, C20:0, C21:0, C23:0, C18:1, C22:1, C22:0, C16:0; preferably from: C23:1, C15:0, C24:2, C20:0, C21:0; more preferably from: C23:1, C15:0, optionally C24:

2.

49. The method according to claim 48, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C23:1, C15:0, C24:2, C20:0, C21:0, C23:0, C18:1, C22:1, C22:0, C16:0; preferably from: C23:1, C15:0, C24:2, C20:0, C21:0, C23:0; more preferably from: C23:1, C15:0, C24:2, C20:0, optionally C21:

0.

50. The method according to any one of claims 3-7, for distinguishing a subject having MG from a subject having PMA wherein the reference SM profiles compared are MG versus PMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C24:2, C22:0, C20:0, C20:1, C15:0; preferably from: C24:2, C22:0, optionally C20:

0.

51. The method according to claim 50, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from:C24:2, C22:0, C20:0, C20:1, optionally C15:

0.

52. The method according to any one of claims 3-7, for distinguishing a subject having MG from a subject having SBMA wherein the reference SM profiles compared are MG versus SBMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C23:1, C21:0, C22:1, C24:2, C23:0, C16:1, C24:0, C18:1, C15:0, C18:0; preferably from: C23:1, C21:0, C22:1, C24:2, C23:0, C16:1; more preferably from: C23:1, C21:0, optionally C22:

1.

53. The method according to claim 52, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C23:1, C21:0, C22:1, C24:2, C23:0, C16:1, C24:0, C18:1, C15:0, C18:0; preferably from: C23:1, C21:0, C22:1, C24:2, C23:0, C16:1, C24:0; more preferably from: C23:1, C21:0, C22:1, C24:2, optionally C23:

0.

54. The method according to any one of claims 3-7, for distinguishing a subject having MG from a subject having SMA wherein the reference SM profiles compared are MG versus SMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C23:1, C23:0, C21:0, C22:1, C24:2, C16:1, C24:0, C15:0, C18:0, C18:1, C20:0; preferably from: C23:1, C23:0, C21:0, C22:1, C24:2, C16:1, C24:0, C15:0; more preferably from: C23:1, C23:0, optionally C21:

0.

55. The method according to claim 54, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C23:1, C23:0, C21:0, C22:1, C24:2, C16:1, C24:0, C15:0, C18:0, C18:1, C20:0; preferably from: C23:1, C23:0, C21:0, C22:1, C24:2, C16:1; C24:0, C15:0; more preferably from: C23:1, C23:0, C21:0, C22:1, optionally C24:

2.

56. The method according to any one of claims 3-7, for distinguishing a subject having PD from a subject having PMA wherein the reference SM profiles compared are PD versus PMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C20:0, C15:0, C23:1, C23:0, C16:0, C22:0; preferably from: C20:0, C15:0, optionally C23:1.

57. The method according to claim 56, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C20:0, C15:0, C23:1, C23:0, C16:0, C22:0; preferably from: C20:0, C15:0, C23:0, C23:1, optionally C16:

0.

58. The method according to any one of claims 3-7, for distinguishing a subject having PD from a subject having SBMA wherein the reference SM profiles compared are PD versus SBMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15:0, C20:0, C22:1, C21:0, C23:0, C22:0, C16:1, C23:1, C24:0, C18:1, C24:1, C24:2, C20:1, C16:0, C18:0; or C15:0, C16:0, C16:1, C18:1, C20:0, C20:1, C21:0, C22:0, C22:1, C23:0, , C24:1, C24:2; preferably from: C15:0, C20:0, C22:1, C21:0, C22:0, C23:0; more preferably from: C15:0, C20:0, optionally C22:

1.

59. The method according to claim 58, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C15:0, C20:0, C22:1, C21:0, C23:0, C22:0, C16:1, C23:1, C24:0, C18:1, C24:1, C24:2, C20:1, C16:0, C18:0; or C15:0, C16:0, C16:1, C18:1, C20:0, C20:1, C21:0, C22:1, C23:0, C24:1, C24:2; preferably from: C15:0, C20:0, C22:1, C21:0, C22:0, C23:0, C16:1; more preferably from: C15:0, C20:0, C22:1, C21:0, optionally C22:

0.

60. The method according to any one of claims 3-7, for distinguishing a subject having PD from a subject having SMA wherein the reference SM profiles compared are PD versus SMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15:0, C20:0, C22.1, C23.1, C23:0, C18:1, C24:0, C21:0, C16:1, C18:0, C16:0, C22:0, C24:1, or C15:0, C16:0, C16:1, C18:0, C18:1, C20:0, C21:0, C22:0, C22:1, C23:1, C24:0; preferably from: C15:0, C20:0, C22.1, C23.1, C23:0, C18:1, C24:0; more preferably from: C15:0, C20:0, optionally C22.

1.

61. The method according to claim 60, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from:C15:0, C20:0, C22.1, C23.1, C23:0, C18:1, C24:0, C21:0, C16:1, C18:0, C16:0, C22:0, C24:1, or C15:0, C16:0, C16:1, C18:0, C18:1, C20:0, C21:0, C22:0, C22:1, C23:1, C24:0; preferably from: C15:0, C20:0, C22.1, C23.1, C23:0, C18:1, C24:0, C21:0, more preferably from: C15:0, C20:0, C22.1, C23.1, optionally C23:

0.

62. The method according to any one of claims 3-7, for distinguishing a subject having PMA from a subject having SBMA wherein the reference SM profiles compared are PMA versus SBMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C21:0, C23:0, C22:0, C22:1, C20:1, C20:0, C24:0, C16:0; preferably from: C21:0, C23:0, C22:0, C22:1, C20:1; more preferably from: C21:0, C23:0, optionally C22:

0.

63. The method according to claim 62, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C21:0, C23:0, C22:0, C22:1, C20:1, C20:0, C24:0, C16:0; preferably from: C21:0, C23:0, C22:0, C22:1, optionally C20:

1.

64. The method according to any one of claims 3-7, for distinguishing a subject having PMA from a subject having SMA wherein the reference SM profiles compared are PMA versus SMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C23:0, C21:0, C18:0, C22:0, C22:1, C24:0, C20:0; preferably from: C23:0, C21:0, C18:0, C22:0; more preferably from: C23:0, C21:0, optionally C18:

0.

65. The method according to claim 64 wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C23:0, C21:0, C18:0, C22:0, C22:1, C24:0, C20:0; preferably from: C23:0, C21:0, C18:0, C22:0, optionally C22:

1.

66. The method according to any one of claims 3-7, for distinguishing a subject having SBMA from a subject having SMA wherein the reference SM profiles compared are SBMA versus SMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C21:0, C22:0, C20:1, C22:1, C24:2, C23:1, C18:0, C18:1, C16:

0. C24:0, C24:1; orC15:0, C18:0, C20:1, C22:0, C24:2; preferably from: C21:0, C22:0, C20:1, C22:1, C24:2, C23:1; more preferably from: C21:0, C22:0, optionally C20:

1.

67. The method according to claim 66, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C21:0, C22:0, C20:1, C22:1, C24:2, C23:1, C18:0, C18:1, C16:

0. C24:0, C24:1; or C15:0, C18:0, C20:1, C22:0, C24:2; preferably from: C21:0, C22:0, C20:1, C22:1, C24:2, C23:1, C18:0; more preferably from: C21:0, C22:0, C20:1, C22:1, optionally C24:

2.

68. The method according to any one of claims 3-7, for distinguishing a subject having SMAII from a subject having SMAIII / IV, wherein the reference SM profiles compared are SMAII versus SMAIII / IV, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15:0, C16:0, C24:0, C18:1, C24:1, C24:2, C18:0, or C15:0, C16:0, C21:0, C22:0, C24:0, C:24:1, C24:2; preferably from: C15:0, C16:0, C24:0, C18:1, C24:1; or preferably from: C15:0, C16:0, C21:0, C22:0, C24:1; more preferably from: C15:0 and / or C16:0, optionally C24:0, or more preferably from: C16:0 and / or C24:0, optionally C15:0, or more preferably from: C24:0 and / or C15:0, optional C16:

0.

69. The method according to claim 66, wherein the “patient” and “reference” SM profile(s) share in common at least four or five key SM observation(s) selected from: C15:0, C16:0, C18:1, C24:1, C18:0, C24:2, C24:0; preferably from: C15:0, C16:0, C24:0, C18:1, C24:1, C24:2; more preferably from: C15:0, C16:0, C24:0, C18:1, optionally C24:1, or more preferably C15:0, C16:0, C18:1, C24:1 optionally C18:0, C24:2, or C24:0..

70. The method according to any one of claims 1-5 or claim 7, for distinguishing a subject having a MND (ALS or SBMA) from a healthy subject, wherein the reference SM profiles compared are ALS, SBMA and healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0;preferably from: C15:0, C16:1, C18:1, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:

0.

71. The method according to claim 70, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0; preferably from: C15:0, C16:1, C18:1, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:

0.

72. A method according to any one of claims 3-5 or 7, for distinguishing a subject having ALS from a subject having SBMA, wherein the reference SM profiles compared are ALS and SBMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0; preferably from: C15:0, C16:1, C18:1, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:

0.

73. The method according to claim 72, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0; preferably from: C15:0, C16:1, C18:1, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:

0.

74. The method according to any one of claims 1-2, claim 5 or claim 7, for distinguishing a subject having ALS from a healthy subject, wherein the reference SM profiles compared are ALS and heathy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C15:0, C16:1, C20.0, C21:0, C22:1, C22:0, C23:1, C23:0, or C16:0, C18:1, C18:0, C20:1, C24:2, C24:1, C24:

0.

75. The method according to claim 74, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from:C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C15:0, C16:1, C20.0, C21:0, C22:1, C22:0, C23:1, C23:0, or C16:0, C18:1, C18:0, C20:1, C24:2, C24:1, C24:

0.

76. A method according to any one of claims 1-4 or claims 6-7, for diagnosing a subject as healthy or as having ALS or SBMA, wherein the reference SM profiles compared are ALS, SBMA and healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0; preferably form C15:0, C16:1, C18:1, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0, or preferably from C16:0 and / or C18:

0.

77. The method according to claim 76, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0; preferably from: C15:0, C16:1, C18:1, C20:1, C20:0, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0, or preferably including: C16:0, C18:

0.

78. The method according to any one of claims 1-4 or claims 6-7, for diagnosing a subject as healthy or as having ALS or SBMA, wherein the reference SM profiles compared are ALS, SBMA and healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C15:0, C16:1, C20.0, C21:0, C22:1, C22:0, C23:1, C23:0, or C16:0, C18:1, C18:0, C20:1, C24:2, C24:1, C24:0, or C16:0, C18:1, C18:0, C20:1, C21:0, C24:2, C24:1, C24:0, or C15:0, C21:0, C22:0, or C15:0, C21:

0.

79. The method according to claim 78, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from:C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C15:0, C16:1, C20.0, C21:0, C22:1, C22:0, C23:1, C23:0, or C16:0, C18:1, C18:0, C20:1, C24:2, C24:1, C24:0, or C16:0, C18:1, C18:0, C20:1, C21:0, C24:2, C24:1, C24:0, or including C15:0, C21:0, C22:0, or including C15:0, C21:

0.

80. The method according to any one of claims 1-4 or claims 6-7, for diagnosing a subject as healthy or as having ALS or SBMA, wherein the reference SM profiles compared are ALS, SBMA and healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C15:0, C16:0, C20.1, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0, or C15:0 and / or C21:0; preferably from: C15.0, C16.0, C20.1, C21.0, C22.1, C22.0, C23.0; more preferably from: C15:0, C21:0, C22:1, C20:1, C23:0; or more preferably C15:0 and / or C21:0, optionally C22:

1.

81. The method according to claim 80, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C15:0, C16:0, C20.1, C21:0, C22:1, C22:0, C23:1, C23:0, C24:2, C24:1, C24:0, or including C15:0 and / or C21:0; preferably from: C15.0, C16.0, C20.1, C21.0, C22.1, C22.0, C23.0; more preferably from: C15:0, C21:0, C22:1, C20:1, optionally C23:

0.

82. The method according to any one of claims 1-2 or claim 5 or claim 7, for diagnosing a subject as healthy or as having PD, wherein the reference SM profiles compared are healthy and PD, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, orC21:0, optionally C23:1, or C21:0, optionally C24:0, or C23:1, optionally C21:0, or C23:1, optionally C24:0, or C24:0, optionally C21:0, or C24:0, optionally C23:

1.

83. The method according to claim 80, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C21:0, optionally C23:1, or C21:0, optionally C24:0, or C23:1, optionally C21:0, or C23:1, optionally C24:0, or C24:0, optionally C21:0, or C24:0, optionally C23:1 84. The method according to any one of claims 1-2 or claim 5 or claim 7, for diagnosing a subject as healthy or as having PD, wherein the reference SM profiles compared are healthy and PD, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1,.C24. or C16:1, C16:0, C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:2, C24:1, C:24.0, or C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C:24.0, or C20:1, C21:0, C23:1, or C20:1 or C21:0, or C23.

1.

85. The method according to claim 84, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1,.C24. or C16:1, C16:0, C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C24:2, C24:1, C:24.0, or C20:1, C20:0, C21:0, C22:0, C23:1, C23:0, C:24.0, or including C20:1, C21:0, C23:1, or C20:1, C21:0, C23:1, C20:0, optionally C24:0, or C20:1, C21:0, C23:1, C20:0, optionally C23:0, or C20:1, C21:0, C23:1, C20:0, optionally C22:

0.

86. A method for diagnosing a subject as healthy or as having a neurological disorder, said method comprising a method according to one or more of claims 1-7, wherein said neurological disorder is any one of ALS, SBMA, SMA or PD.

87. The method according to claim 86 when dependent on any one or more of claims 1-4 or claims 6-7, wherein the reference SM profiles compared are ALS, SBMA, SMA, PD and healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0; preferably from: C15:0, C24:2, C21:0, C22:0, C24:1, or preferably from: C15:0, C16:0, C18:1, C24:1, C18:0, C24:2, C24:0, or .

88. The method according to claim 86, when dependent on any one or more of claims 1-4 or claims 6-7, wherein said neurological disorder is one of ALS or a Disease Control (SBMA, SMA and PD combined), and wherein the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0; or C15:0, C16:1, C20.0, C21:0, C22:1, C22:0, C23:1, C23:0, or C16:0, C18:1, C18:0, C20:1, C24:2, C24:1, C24:

0.

89. The method according to claim 86, when dependent on any one of more of claims 1-5 or claims 6-7, wherein said neurological disorder is a Disease Control (SBMA, SMA and PD combined) and ALS, wherein the reference SM profiles compared are ALS, SBMA, SMA, PD and healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0, or C15:0, C16:1, C20.0, C21:0, C22:1, C22:0, C23:1, C23:0, or C16:0, C18:1, C18:0, C20:1, C24:2, C24:1, C24:

0.

90. A method for diagnosing a subject as healthy or as having a neurological disorder, said method comprising a method according to any one or more of claims 1-7, wherein said neurological disorder is any one of ALS, SBMA, SMA, PD or AD.

91. The method according to claim 90 when dependent on any one or more of claims 1-5 or claim 7, wherein the reference SM profiles compared are ALS, SBMA, SMA, PD, AD and healthy, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from:C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0; preferably from: C15:0, C20:1, C21:0, C22:1, C22:0, C16:0, C23:1, C24:

0. more preferably from: C15:0, C20:1, optionally C21:

0.

92. The method according to claim 91, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: C15.0, C16.1, C16.0, C18.1, C18.0, C20.1, C20.0, C21.0, C22.1, C22.0, C23.1, C23.0; C24:2, C24:1, C24.0; preferably from: C15:0, C20:1, C21:0, C22:1, C22:0, C16:0, C23.1; more preferably from: C15:0, C20:1, C21:0, C22:1, optionally C22:0, C16:

0.

93. The method according to any one of claims 3-7, for distinguishing a subject having ALS (non-PMA / PLS) from a subject having PLS, wherein the reference SM profiles compared are ALS (non PMA / PLS) versus PLS, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C23.1, C22:1, C24:1, C21:0, C16:0; preferably from: C23:1 and / or C22:1, optionally C24:1 94. The method according to claim 93, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: C23.1, C22:1, C24:1, C21:0, optionally C16:

0.

95. The method according to any one of claims 3-7, for distinguishing a subject having ALS (non-PMA / PLS) from a subject having PMA, wherein the reference SM profiles compared are ALS (non-PMA / PLS) v PMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C23.0, C23:1, C18:0, C22:0, C16:0; preferably from: C16:0, C22:0, C23:0; more preferably from: C16:0 and / or C22:0, optionally C23:0, or more preferably from:C22:0 and / or C23:0, optionally C16:0, or more preferably from: C23:0 and / or C16:0, optionally C22:

0.

96. The method according to claim 95, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: Including: C23.0, C23:1, C18:0, C22:0, C16:0.

97. The method according to any one of claims 3-7, for distinguishing a subject having PLS from a subject having PMA, wherein the reference SM profiles compared are PLS v PMA, and the “patient” and “reference” SM profile(s) share in common one or more SM observation(s) selected from: C23:1, C24:1, C23:0, C18:0, C20:1; preferably from: C23:1 and / or C24:1, optionally C23:0, or preferably from:C24:1 and / or C23:0, optionally C23:1, or preferably from: C23:1 and / or C24:1, optionally C23:

0.

98. The method according to claim 97, wherein the “patient” and “reference” SM profile(s) share in common at least four or five SM observation(s) selected from: C23.1, C24:1, C23:0, C18:0, optionally C20:

1.

99. A method for diagnosing a neurological disorder in a subject, wherein the subject presents asymptomatically with regard to said neurological disorder, comprising a method according to any one or more of the preceding claims.

100. A method for screening a subject to ensure said subject has a first neurological disorder, wherein said first neurological disorder defines the correct neurological disorder qualifying criteria for enrolment into a clinical trial, said method comprising a method according to any one or more of the preceding claims, wherein said method confirms the subject has the first neurological disorder.

101. The method according to claim 100, said method comprising a method according to any one or more of claims 3-7 and / or any one or more of claims 22-100 to determine whether said subject has a second neurological disorder, and confirming the subject does not have said second neurological disorder.

102. A method for monitoring responsiveness to therapeutic intervention (e.g. gene or gene expression modifying therapy) in a subject having a first neurological condition, said method comprising: a) providing a patient primary sphingomyelin (SM) profile for said subject, i) wherein said patient primary SM profile is obtainable by mass spectrometry analysis of a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject at a selected timepoint during or after therapeutic intervention; andii) wherein said patient primary SM profile comprises one or more primary SM observation(s), each of which is a different SM observation(s); b) providing a reference primary SM profile comprising one or more primary SM observations: i) wherein said reference primary SM profile is obtainable by comparative mass spectrometry analysis of body fluid samples comprising 30-300nm EVs isolated from two different reference cohorts selected from a healthy subject cohort and a cohort of subjects having the first neurological disorder; ii) wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD) value for said one or more primary SM observation(s); and (b) between each reference cohort, identification of one or more key SM observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and iii) wherein said reference primary SM profile comprising one or more of said key SM observation(s) is provided in a form that is suitable for comparative analysis with the patient primary SM profile; c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that: i) when the healthy subject cohort is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has been successful, whereas a “non-match” is indicative the subject has the first neurological disorder, and wherein when said “degree of non-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention,an observed trend away from a “non-match” status or towards a “match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention; or ii) when the cohort of subjects having the first neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has not been successful, whereas a “non-match” is indicative the subject has the first neurological disorder, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “match” status or towards a “non-match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention.

103. The method according to claim 102, for monitoring responsiveness to therapeutic intervention (e.g. gene or gene expression modifying therapy) of a subject having either SMA type II disorder or SMA type III / IV disorder, wherein the first neurological disorder is either SMA type II or SMA type III / IV, and wherein step c) of said method comprises: c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that:i) when the healthy subject cohort is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has been successful, whereas a “non-match” is indicative the subject has the first neurological disorder, and wherein when said “degree of non-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “non-match” status or towards a “match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention; or ii) when the cohort of subjects having the first neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has not been successful, whereas a “non-match” is indicative the subject has the first neurological disorder, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), orcompared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “match” status or towards a “non-match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention.

104. The method according to claim 102, for monitoring responsiveness to therapeutic intervention (e.g. gene or gene expression modifying therapy) of a subject having either Fast-progression ALS or Slow-progression ALS, wherein the first neurological disorder is either Fast-progression ALS or Slow-progression ALS, and wherein step c) of said method comprises: c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that: i) when the healthy subject cohort is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has been successful, whereas a “non-match” is indicative the subject has the first neurological disorder, and wherein when said “degree of non-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or comparted with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “non-match” status or towards a “match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observedtrend(s) is indicative the subject has derived no benefit from the therapeutic intervention or ii) when the cohort of subjects having the first neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has not been successful, whereas a “non-match” is indicative the subject has the first neurological disorder, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “match” status or towards a “non- match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention.

105. The method according to claim 102, for monitoring responsiveness to therapeutic intervention (e.g. gene or gene expression modifying therapy) of a subject having either Severe ALS or Mild ALS, wherein the first neurological disorder is either Severe ALS or Mild ALS, and wherein step c) of said method comprises: c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that: i. when the healthy subject cohort is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SMobservation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has been successful, whereas a “non-match” is indicative the subject has Severe ALS, and wherein when said “degree of non-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “non-match” status or towards a “match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention; or ii. when the cohort of subjects having Severe ALS is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, wherein a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has not been successful, whereas a “non-match” is indicative the subject has Severe ALS, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “match” status or towards a “non- match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observedtrend(s) is indicative the subject has derived no benefit from the therapeutic intervention.

106. A method for monitoring responsiveness to therapeutic intervention (e.g. gene or gene expression modifying therapy) in a subject having a first neurological condition, said method comprising: a) providing a patient primary sphingomyelin (SM) profile for said subject, i. wherein said patient primary SM profile is obtainable by mass spectrometry analysis of a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject at a selected timepoint during or after therapeutic intervention, and ii wherein said patient primary SM profile comprises one or more primary SM observation(s), each of which is a different SM observation(s); b) providing a reference primary SM profile comprising one or more primary SM observations:, i. wherein said reference primary SM profile is obtainable by comparative mass spectrometry analysis of body fluid samples comprising 30-300nm EVs isolated from two different reference cohorts selected from a cohort of subjects having the second neurological disorder and a cohort of subjects having the first neurological disorder, and wherein the second disorder is a less advance or less aggressive form of the first neurological disorder; ii. wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD) value for said one or more primary SM observation(s); and (b) between each reference cohort, identification of one or more key SM observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and iii. wherein said reference primary SM profile comprising one or more of said key SM observation(s) is provided in a form that is suitable for comparative analysis with the patient primary SM profile; c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that: i. when the cohort of subject having the second neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s)compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has been successful, whereas a “non-match” is indicative the subject has the first neurological disorder, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “match” status or towards a “non- match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention; or ii. when the cohort of subjects having the first neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, wherein a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has not been successful, whereas a “non- match” is indicative the subject has the first neurological disorder, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention,an observed trend away from a “match” status or towards a “non- match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention.

107. The method according to claim 106, for monitoring responsiveness to therapeutic intervention (e.g. gene or gene expression modifying therapy) of a subject having a SMA type II disorder, wherein the first neurological disorder is SMA type II and the second neurological disorder is SMA type III / IV, and wherein step c) of said method comprises: c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that i) when the cohort of subjects having SMA type III / IV is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has been successful in that the patient has regressed to SMA type III / IV, whereas a “non-match” is indicative the subject has SMA type II, and wherein when said “degree of non-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “non-match” status or towards a “match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention; orii) when the cohort of subjects having SMA type II is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, wherein a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has not been successful, whereas a “non-match” is indicative the subject has SMA type II, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “match” status or towards a “non- match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention.

108. The method according to claim 106, for monitoring responsiveness to therapeutic intervention (e.g. gene or gene expression modifying therapy) of a subject having Fast progression ALS, wherein the first neurological disorder is Fast progression ALS and the second neurological disorder is Slow-progression ALS, and wherein step c) of said method comprises: c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that i) when the cohort of subjects having Slow progression ALS is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile fallswithin 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has been successful in that the patient has regressed to Slow progression ALS, whereas a “non-match” is indicative the subject has Fast progression ALS, and wherein when said “degree of non-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “non-match” status or towards a “match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention; or ii) when the cohort of subjects having Fast progression ALS is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, wherein a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has not been successful, whereas a “non-match” is indicative the subject has Fast progression ALS, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “match” status or towards a “non- match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observedtrend(s) is indicative the subject has derived no benefit from the therapeutic intervention.

109. The method according to claim 106, for monitoring responsiveness to therapeutic intervention (e.g. gene or gene expression modifying therapy) of a subject having Severe ALS, wherein the first neurological disorder is Severe ALS and the second neurological disorder is Mild ALS, and wherein step c) of said method comprises: c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that i) when the cohort of subjects having Mild ALS is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has been successful in that the patient has regressed to Mild ALS, whereas a “non-match” is indicative the subject has Severe ALS, and wherein when said “degree of non-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “non-match” status or towards a “match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention; or ii) when the cohort of subjects having Severe ALS is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) forone or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the therapeutic intervention has not been successful, whereas a “non-match” is indicative the subject has Severe ALS, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same time period of therapeutic intervention, an observed trend away from a “match” status or towards a “non- match” status that correlates with an increasing time period of therapeutic intervention is indicative the subject has derived benefit from the therapeutic intervention, whereas the absence of said observed trend(s) is indicative the subject has derived no benefit from the therapeutic intervention.

110. A method for monitoring progression of a first neurological disorder, said method comprising a) providing a patient primary sphingomyelin (SM) profile for said subject, i. wherein said patient primary SM profile is obtainable by mass spectrometry analysis of a body fluid sample comprising 30-300nm extracellular vesicles (EVs) isolated from the subject at a selected timepoint during or after therapeutic intervention, and ii. wherein said patient primary SM profile comprises one or more primary SM observation(s), each of which is a different SM observation(s); b) providing a reference primary SM profile comprising one or more primary SM observations:, i) wherein said reference primary SM profile is obtainable by comparative mass spectrometry analysis of body fluid samples comprising 30-300nm EVs isolated from two different reference cohorts selected from a cohort of subjects having the second neurological disorder and a cohort of subjects having the first neurological disorder, and wherein the second disorder is a more advance or more aggressive form of the first neurological disorder;ii) wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD) value for said one or more primary SM observation(s); and (b) between each reference cohort, identification of one or more key SM observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and iii) wherein said reference primary SM profile comprising one or more of said key SM observation(s) is provided in a form that is suitable for comparative analysis with the patient primary SM profile; c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that: i. when the cohort of subjects having the second neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the patient’s disorder has transitioned to the second neurological disorder, whereas a “no-match” is indicative the subject has not transitioned to the second neurological disorder, and wherein when said “degree of match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same passage of time, an observed trend towards a “match” status or away from a “non- match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to that of the second neurological disorder than it is to that of the first neurological disorder, whereas an observed trend away from a “match” status or towards a “non-match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to that of the firstneurological disorder than it is to that of the second neurological disorder; or ii. when the cohort of subjects having the first neurological disorder is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the patient’s disease stage has not transitioned beyond the first neurological disorder (at the timepoint the patient sample was isolated), whereas a ”no-match” is indicative the patient’s disease state has transitioned beyond the first neurological disorder, and wherein when said “degree of no-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same passage of time, an observed trend towards a “match” status or away from “no- match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to that of the first neurological disorder than it is to that of the second neurological disorder, whereas an observed trend away from a “match” status or towards a “non-match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to that of the second neurological disorder than it is to that of the first neurological disorder..

111. The method according to claim 110, for monitoring progression of Slow- progression ALS to Fast-progression ALS, wherein the first neurological disorder is Slow progression ALS and the second neurological disorder is Fast progression ALS, and wherein step c) of said method comprises: c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that:i. when the cohort of subjects having Fast progression ALS is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the patient’s disease stage has transitioned to Fast progression ALS, whereas a “no-match” is indicative the subject’s disease stage has not transitioned to Fast progressive ALS, and wherein: when said “degree of no-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same passage of time, an observed trend towards a “match” status or away from a “no- match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to Fast progressive ALS than it is to that to Slow progressive ALS, whereas an observed trend away from a “match” status or towards a “non-match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to Slow progressive ALS than it is to that of Fast Progressive ALS; or ii. when the cohort of subjects having Slow progression ALS is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the patient’s disease stage has not progressed beyond Slow progression ALS, whereas a ”no-match” is indicative the subject’s disease stage has transitioned beyond Slow progression ALS, and whereinwhen said “degree of no-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same passage of time, an observed trend towards a “match” status or away from a “no- match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to Slow progressive ALS than it is to that to Fast progressive ALS, whereas an observed trend away from a “match” status or towards a “non-match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to Fast progressive ALS than it is to that of Slow Progressive ALS.

112. The method according to claim 110, for monitoring progression of Mild ALS to Severe ALS, wherein the first neurological disorder is Mild ALS and the second neurological disorder is Severe ALS, and wherein step c) of said method comprises: c) comparing one or more of said key SM observation(s) present in the reference primary SM profile with the corresponding SM observation(s) present in the patient primary SM profile, such that: i. when the cohort of subjects having Severe ALS is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the patient’s disease stage has transitioned to Fast progression ALS, whereas a “no-match” is indicative the subject’s disease stage has not transitioned to Severe ALS, and wherein: when said “degree of no-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or compared with an equivalent control value(s) calibrated to reflect the same passage of time,an observed trend towards a “match” status or away from a “no- match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to Severe ALS than it is to that to Mild ALS, whereas an observed trend away from a “match” status or towards a “non-match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to Mild ALS than it is to that of Severe ALS; or ii. when the cohort of subjects having Mild ALS is selected as the reference primary SM profile and the mean-fold expression value(s) for one or more of said key SM observation(s) compared with the corresponding expression value(s) for the same SM observation(s) of the patient primary SM profile, a “match” is defined when the value(s) for one or more key SM observation(s) of the patient primary SM profile falls within 2 SD of the mean expression value(s) for said corresponding one of more key SM observation(s) of the reference primary SM profile, and wherein a “match” is indicative the patient’s disease stage has not progressed beyond Slow progression ALS, whereas a ”no-match” is indicative the subject’s disease stage has transitioned beyond Slow progression ALS, and wherein when said “degree of no-match” is compared with one or more otherwise identical calculation(s) performed on one or more otherwise identical patient sample(s) isolated at a different timepoint(s), or comparted with an equivalent control value(s) calibrated to reflect the same passage of time, an observed trend towards a “match” status or away from a “no- match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to Mild ALS than it is to that to Severe ALS, whereas an observed trend away from a “match” status or towards a “non-match” status that correlates with an increasing passage of time is indicative the patient’s disease state is closer to Severe ALS than it is to that of Mild ALS.

113. The method according to claim 107, said method comprising the method of claim 68 or claim 69.

114. The method according to claim 108 or claim 111, said method comprising the method of claim 22 or claim 23.

115. The method according to claim 109 or claim 112, said method comprising the method of claim 24 or claim 25.

Images