A method of diagnosing and / or differentiating a neurological disorder

Abstract

The present invention relates to a method for diagnosing a patient having a neurological (including neuromuscular) disorder, such as ALS. 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.

Description

[0001] A method of diagnosing and / or differentiating a neurological disorder

[0002] Field of the invention

[0003] 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 atrophy type II (astasia) from spinal muscular atrophy type III or type IV (mobile but may subsequently lose ambulation); “fastprogression” 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.

[0004] Background to the invention

[0005] 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.

[0006] 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.

[0007] 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.

[0008] 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.

[0009] 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.

[0010] 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.

[0011] 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.

[0012] 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.

[0013] 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.

[0014] Notwithstanding the above-noted short comings, neurofilament protein remains the current Gold Standard biomarker for the diagnosis of neurological diseases.

[0015] 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

[0016] 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..

[0017] In more detail, the present invention flows from the quantitative and qualitative analysis of key lipid 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 lipid 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 lipid observations relied on.

[0018] Data for lipids is typically obtained by mass spectrometry, most preferably liquid chromatography mass spectrometry. LC-MS can provide absolute values for every compound / lipid, using an unbiased approach measuring all lipids extracted from exosomes.

[0019] According to a first aspect of the present invention, there is provided a method for diagnosing a neurological disorder in a subject, said method comprising

[0020] a) providing a patient lipid profile for said subject,

[0021] i. wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, and

[0022] ii. wherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;

[0023] b) providing a reference lipid profile comprising one or more lipid observation(s):

[0024] i. wherein said reference lipid 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 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 lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and

[0025] iii. wherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;

[0026] c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:

[0027] i. when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has the neurological disorder; and / or

[0028] ii. when the cohort of subjects having the neurological disorder is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject has the neurological disorder and / or wherein a “non-match” is indicative the subject is healthy;

[0029] d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said top 5 rankedobservations is employed at any one time, said 5 top-ranked observations being selected from:

[0030] i.as a healthy reference observation:

[0031] SM 38:2; and

[0032] ii.as a neurological disorder observation:

[0033] SM 36:2, PC 34:1, PC 36:1 & SM 34:2

[0034] wherein a “match” with said healthy reference observation and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthy

[0035] whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

[0036] The neurological disorder is most preferably a neurodegenerative disorder.

[0037] The skilled person understands the terminology used to describe these lipids. For example, the skilled person would understand that SM is a sphingomyelin, and details on given lipids can be found in online databases such as https: / / www.metabolomicsworkbench.org / databases / refmet / refmet_details.php7REFMET_ NAM E=SM+38%3A1 %3BO2

[0038] Such first aspect allows for ‘generic’ diagnosis of the patient as having a neurological disorder (e.g. even if the precise type of disorder is not yet specifically diagnosed). Preferably, the first aspect if for diagnosing whether the patient has any one of ALS, PD or AD (e.g. any one of the three). That It may be said that a patient diagnosed with a disorder under the first aspect may have a stronger likelihood to have ALS than any other neurological disorder, or may have a stronger likelihood to have ALS than either of PD or AD.

[0039] A second aspect provides a method for diagnosing amyotrophic lateral sclerosis (ALS) disorder in a subject, said method comprising

[0040] a) providing a patient lipid profile for said subject,

[0041] i) wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, andii) wherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;

[0042] b) providing a reference lipid profile comprising one or more lipid observation(s):

[0043] i) wherein said reference lipid 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 ALS;

[0044] 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 lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and

[0045] iii) wherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;

[0046] c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:

[0047] i) when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has ALS; and / or

[0048] ii) when the cohort of subjects having ALS is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipidobservation(s) of the reference lipid profile, said “match” being indicative the subject has ALS and / or wherein a “non-match” is indicative the subject is healthy;

[0049] d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said 5 top-ranked observations is employed at any one time, said 5 top-ranked observations being selected from:

[0050] i) as an ALS observation:

[0051] SM 36:2, PC 34:1, SM 34:2, PC 0-34:1, & LPC 20:4;

[0052] wherein a “match” with said healthy reference observation and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthy,

[0053] whereas a “non-match” with said healthy reference observation and “match” with said ALS reference observations is indicative the subject has ALS.

[0054] A preferred aspect of the invention provides a method for diagnosing Amyotrophic lateral sclerosis (ALS) in a subject, said method comprising

[0055] a) providing a patient lipid profile for said subject,

[0056] i. wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, and

[0057] ii. wherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;

[0058] b) providing a reference lipid profile comprising one or more lipid observation(s):

[0059] i. wherein said reference lipid 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 ALS;

[0060] 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 lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s)that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and

[0061] iii. wherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;

[0062] c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:

[0063] i. when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has ALS; and / or

[0064] ii. when the cohort of subjects having ALS is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject has ALS and / or wherein a “non-match” is indicative the subject is healthy;

[0065] d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said top 5 ranked observations is employed at any one time, said 5 top-ranked observations being selected from:

[0066] as an ALS observation:

[0067] PC 0-34:1, PC 34:1 and SM36:2;

[0068] as a healthy reference observation:SM 41:1 and PC 32:1;

[0069] wherein a “match” with said healthy reference observation and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthy

[0070] whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

[0071] Another aspect provides method for diagnosing Alzheimer’s Disease (AD) in a subject, said method comprising

[0072] a) providing a patient lipid profile for said subject,

[0073] wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, and

[0074] wherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;

[0075] b) providing a reference lipid profile comprising one or more lipid observation(s):

[0076] wherein said reference lipid 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 AD; 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 lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and

[0077] wherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;

[0078] c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:

[0079] when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said keylipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has the AD; and / or

[0080] when the cohort of subjects having the AD is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject has the AD and / or wherein a “non-match” is indicative the subject is healthy;

[0081] d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said 5 top-ranked observations is employed at any one time, said 5 top-ranked observations being selected from:

[0082] as a healthy reference observation:

[0083] SM 38:2, PC36:2, & PC40:4; and

[0084] as an AD observation:

[0085] SM34:2, & PC40:5;

[0086] wherein a “match” with said healthy reference observation and a “non- match” with said AD reference observations is indicative the subject is healthy,

[0087] whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD.

[0088] Another aspect provides a method for diagnosing Parkinson’s Disease (PD) in a subject, said method comprisinga) providing a patient lipid profile for said subject,

[0089] wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, and

[0090] wherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;

[0091] b) providing a reference lipid profile comprising one or more lipid observation(s):

[0092] wherein said reference lipid 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 PD;

[0093] 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 lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; and

[0094] wherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;

[0095] c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:

[0096] when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has the PD; and / or

[0097] when the cohort of subjects having the PD is selected as the reference lipid profile and the mean-fold expression value(s) for one ormore of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject has the PD and / or wherein a “non-match” is indicative the subject is healthy;

[0098] d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said 5 topranked observations is employed at any one time, said 5 top-ranked observations being selected from:

[0099] as a healthy reference observation:

[0100] PC 34:4, PC 36:2, PC 40:4, SM 32:2; and

[0101] as a PD observation:

[0102] SM 40:2;

[0103] wherein a “match” with said healthy reference observation and a “nonmatch” with said PD reference observations is indicative the subject is healthy,

[0104] whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD.

[0105] Throughout this specification, it is preferred that a sample (e.g. patient sample) is subjected to mass spectrometry to provide a patient lipid profile for said subject of step a). Thus, the patient lipid profile is preferably obtained by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject.

[0106] The skilled person would understand that the lipids described herein are analysable as ionic species / adduct ions e.g. as [M + NH4]+ adduct ions, [M + H]+ adduct ions, or [M + Na]+ adduct ions. The preferred adducts for lipids described herein are outlined in the figures. Figure 14C also shows the preferred adduct for each key lipid observation for ALS (e.g. the preferred adduct for PC 0-34:1 is [M+H]+). Figure 15D also shows thepreferred adduct for each key lipid observation for SMA (e.g. the preferred adduct for PC SM 42:2 is [M+H]+).

[0107] Detailed description of the invention

[0108] 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 lll / 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).

[0109] 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.

[0110] Each reference lipid profile comprises one or more lipid observation(s) and a comparison of the same one or more lipid observation(s) when shared in common between two reference SM profiles allows one to select one or more key lipid observation(s) that demonstrates a statistically significant difference, thereby providing statistical basis upon which the two reference lipid profiles are distinguishable. The patient lipid profile is then compared with either (or both) reference lipid profile(s) and the relative expression level for said one or more key lipid observation(s) interrogated. Finally, since the observed expression level for each key lipid observation(s) of the patient lipid profile will match more closely to the calculated mean expression level for one reference lipid 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.

[0111] The lipid chemical structures listed above are considered the conventional lipid (e.g. 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, a lipid (e.g. SM) variant form may be considered a unique lipid (e.g. SM) structure in its own right or subsumed as a conventional isomer (according to its structural “best match”).Thus, to the extent a lipid (e.g. SM) variant is identified, it is preferable that the lipid (e.g. SM) variant be treated consistently (i.e. as a unique lipid e.g. SM, or a conventional isomer) when preparing all lipid (e.g. SM) profiles (i.e. both lipid e.g. SM reference profiles and lipid e.g. SM patient profile) for use in the same diagnostic method of the present invention.

[0112] The present invention relies on analysis of the lipid 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 the sub-populations of lipids described here is / are (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 lipid expression profiles for the EV-isolated sub-population of lipids were significantly more robust than any corresponding lipid 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.

[0113] 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”).

[0114] When considering which lipid observation(s) to select as a key lipid observation(s) for use in a diagnostic assay of the present invention, one would typically start by identifying candidate lipid 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 lipid observation(s) required to create a reference lipid profile, one may prefer to select the lipid observation that demonstrates the highest + / - fold-change value, and then proceed toselect 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 three lipid observations (the three highest fold-change lipid observations), four lipid observations (the four highest fold-change lipid observations), and optionally five lipid observations (the five highest fold-change lipid 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 lipid profile” selection criteria. Accordingly, each reference cohort includes (for each lipid observation) a statistical mean lipid expression value. This mean cohort lipid expression value is compared with a corresponding mean cohort expression value for the same lipid 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.

[0115] So, applying the above principles to the present invention, the mean expression value for each lipid observation in a first reference profile (for a first neurological disorder) is compared with the corresponding mean expression level for the same lipid observation in a second reference profile (for a different neurological disorder or a healthy control).

[0116] Next, one can apply 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 lipid observations having a p-value of 0.05 or more.

[0117] Having decided on an appropriate p-value (e.g. p <0.05) for the diagnostic method in question, one can then apply an appropriate threshold criterion in order to provide a final selection of key lipid observations, any number of which may be chosen for inclusion in a “reference lipid profile” of choice (for each of the considered diagnostic assays). As briefly discussed earlier, one may choose to select say the top three or four or five lipid 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-changethreshold 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

[0118] It is preferred that the lipids chosen for analysis are based on the value of their associated feature weights (importance) in a prediction model e.g. trained on lipid profile data from both healthy and disease patients. Feature importance is used in machine learning to understand which features in a dataset contribute the most to a model's predictions, allowing one to identify the most relevant features, improve model interpretability, make informed decisions about feature selection, and focus on the most impactful variables.

[0119] This goes toward explaining the terms “wherein the key lipid observations are provided in the form of a ranked list of lipid observations” and “the 5 top-ranked key observations”. As described in Example 2 (“Feature importance” section), a model was trained on 51 lipid observations obtained from LC-MS analysis of EV lipid levels from both healthy and disease cohorts, and the importance (e.g. “feature weight) of any given lipid for pushing toward diagnosis of healthy or disease state was determined. The 51 lipids could then be ‘ranked’ by importance according to the ranking of “feature weights table” and / or figure 7. Thus, the first 5 lipids in the table or figure list (with the highest positive or negative feature values) represent “the 5 top-ranked key observations”.

[0120] More information on the extracellular vesicles will now be provided. 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.

[0121] 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, orgreater 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),

[0122] 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).

[0123] 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).

[0124] Different methods can be used to extract EVs, including: differential ultracentrifugation, polymer-based precipitation, size exclusion chromatography, ultrafiltration, and / or immunoaffinity.

[0125] To extract lipid from the EVs for mass spectrometry analysis, exosome samples can be mixed with internal standard (e.g. SPLASH standard) together with 400-800pl of chloroform:methanol, which may be at a ratio or 10:1, 8:1, 6:1, 4:1 or 2:1. Water (e.g. 100-300 pl) can then be added, followed by centrifugation at 13,000g and collection and transferral of the lipid fraction into a new tube. Samples can then be dried by to evaporate solvent (e.g. SpeedVac), following by reconstitution in 25-100pl of chloroform: methanol (at a ratio or 10:1, 8:1, 6:1, 4:1 or 2:1). Reconstituted samples can then be subjected to analysis or frozen for future use.

[0126] 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 anoutput is provided in the form of a predicted probability score that the subject has or does not have the neurological disorder in question.

[0127] 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.

[0128] Brief description of the drawings

[0129] The present invention is described with reference to the appended non-limiting examples and the accompanying drawings, in which:

[0130] Figure 1 (a)-(C) illustrates results of principal components analysis of over 100 lipid candidate biomarkers (extracted from extracellular vesicles from sera) and shows global changes in lipid levels in disease vs healthy state subjects, as described in Example 1. (E) - confusion matrices after scaling data separately and using a single predictive model, with a combined accuracy of 85.11 (+ / - 2.22 std) %. Here, we scale the data from the boxes separately and apply a logistic regression classifier to predict whether a sample is labelled as D7 or D1. The classifier trains and predicts on the data using a cross-validation strategy, which is run for 100 repeats. The figure shows the mean confusion matrix for these repeats. D1 = ALS; D7 = healthy control.

[0131] Figure 2 (a) illustrates feature sensitivity plots distinguishing the sensitivity / accuracy in relation to the number of patients in the respective cohort. (B) - illustrates predicted probability plots of 40 features when comparing two disease conditions.

[0132] Figures 3 provides an illustration of the 5-fold cross validation employed during the machine learning analysis described in Example 2.

[0133] Figure 4 provides graphs for PCA before and after batch scaling.

[0134] Figures 5 provides a graph demonstrating logistic regression feature weights.

[0135] Figures 6 shows classification metrics as the number of features increases.

[0136] Figure 7 illustrates the feature importance / weight rankings (for 51 lipids) when comparing individual diseases vs healthy patients.Figure 8 - probability of disease after treatment. The probability of the samples being labelled as class D1 at different time points. TO corresponds to a time before treatment and T6 and T12 correspond to the sample derived from a patient being treated for 6 months or 12 months respectively. The model was trained on one box and tested on the other, and the experiment was run.

[0137] Figure 9 illustrates a separation gradient from Liquid chromatography mass spectrometry (LC-MS) analysis. Separation was established by a 18 minute gradient starting from 32% mobile phase A (ACN:H2O 60:40 (v / v), 10 mM AF) to 97% mobile phase B (IPA:ACN 90:10 (v / v), 10 mM AF) at a flow rate of 0.25 ml / min.

[0138] Figure 10 - Confusion matrix for the Logistic Regression model classifying ALS vs control. The model was trained on data from 44 controls and 112 ALS samples using balanced class weighting. Values represent the number of correctly and incorrectly classified samples for each true class.

[0139] Figure 11 - Feature importance and model interpretability for ALS vs control classification, (a) Displays model coefficients for the most influential features in the Logistic Regression model. Positive coefficients indicate features associated with an increased likelihood of ALS classification, whereas negative coefficients indicate association with controls, (b) Presents SHapley Additive exPlanations (SHAP) values illustrating the direction and magnitude of each feature’s contribution to the model’s predictions across samples, with colour indicating relative feature value. Together, these plots highlight the key lipid biomarkers and demographic variables driving model performance and provide insight into the factors underlying ALS discrimination. Gender is coded as follows: 0-Male, 1 -Female. Age2refers to the squared value of age (normalised by mean) to directly indicate the nonlinearity of the age contribution.

[0140] Figure 12 - Comparison of age and sex distributions before and after demographic imputation. Imputation was performed using biomarker data alone and did not incorporate diagnostic labels. The resulting imputed dataset achieves a more balanced age and sex representation across classes while preserving the overall demographic distributions observed in the original sample.

[0141] Figure 13 - Confusion matrix for the Logistic Regression model classifying ALS vs control using imputed data. The model was trained on data from 166 controls and 112 ALSsamples using balanced class weighting. Values represent the number of correctly and incorrectly classified samples for each true class.

[0142] Figure 14- Feature importance and model interpretability for ALS vs control classification using imputed data, (a) Displays model coefficients for the most influential features in the Logistic Regression model. Positive coefficients indicate features associated with an increased likelihood of ALS classification, whereas negative coefficients indicate association with controls, (b) Presents SHapley Additive exPlanations (SHAP) values illustrating the direction and magnitude of each feature’s contribution to the model’s predictions across samples, with colour indicating relative feature value. Together, these plots highlight the key lipid biomarkers and demographic variables driving model performance and provide insight into the factors underlying ALS discrimination. Gender is coded as follows: O-Male, 1-Female. Age2refers to the squared value of age (normalised by mean) to directly indicate the nonlinearity of the age contribution. The models in A and B were trained with duplicates, the plots in C and D are for models trained without duplicates. Both trained on the same dataset with age and sex imputed.

[0143] Figure 15 - confusion matrix (A), Logistic Regression weights (B), and SHAP plot (C) for the SMA model. The models above were trained with duplicates, the plots in D and E are for models trained without duplicates. Both trained on the same dataset.

[0144] Examples

[0145] Materials and Methods

[0146] EV extraction:

[0147] EV extraction from serum samples

[0148] Different methods are available for isolating exosomes (reviewed in: Sidhom K, et al, Int J Mol Sci 2020). In short, venous blood is collected in sodium citrate tubes from patients and controls and platelet-free plasma is centrifuged at 20,000g for 1 h. The pellet is washed in 0.2 pm filter filtered 1X PBS (Sigma-Aldrich®). The supernatant of exosomes is filtered through a 0.2 pm filter and spun at 100,000 g for 1h at 4 °C and pellet is washed with 1 mL of filtered PBS.

[0149] Alternatively, serum aliquots are thawed on ice and centrifuged at 2,000g for 30 min. The supernatant is transferred into a tube. An aliquot of Total Exosome Isolation (Life Technologies®) is added as per manufacturer’s prescription. The sample is vortexed andincubated for 30min. Subsequently the sample is centrifuged at 10,000g for 10 min, then the supernatant removed. 50 ul of PBS is added. The sample can then be stored at 4 degrees.

[0150] Extraction of lipids

[0151] Exosomes / EVs from were thawed on ice. 25 pl of exosomes / EVs were added to 25pl of SPLASH standard adding 600pl of chloroform:methanol. Tubes were vortexed, 200 pl of water was added, and tubes were vortexed again.

[0152] Samples were then centrifuged at 13,000g. The lipid fraction was collected and transferred into a new tube. Samples were dried by SpeedVac, 50pl of methanokchloroform was added, and samples are stored frozen until use.

[0153] Liquid chromatography mass spectrometry (LC-MS) analysis - Lipidomics Analysis [PDA] S LC-MS mass spectroscopy was used to analyse lipid levels extracted from exosomes / EVs. LC-MS provides absolute values for every compound and 2) using an unbiased approach measuring all lipids extracted from exosomes.

[0154] LC-MS lipidomics analysis was adapted from Breitkopf et al. (Metabolomics. 2017 Mar;13(3):30). In brief, lipid separation was performed on a Thermo Scientific Vanquish HPLC system equipped with a Thermo Scientific Hypersil Gold C18 analytical column (10 cm, ID 2.1 mm, 1.9 pM). Separation was established by a 18 minute gradient starting from 32% mobile phase A (ACN:H2O 60:40 (v / v), 10 mM AF) to 97% mobile phase B (IPA:ACN 90:10 (v / v), 10 mM AF) at a flow rate of 0.25 ml / min. The gradient is shown in figure 9.

[0155] The HPLC system was coupled to an Exploris 480 mass spectrometer (Thermo Scientific). The MS was operated in Data Dependent Acquisition (DDA) mode. DDA was performed with alternating positive negative / ionization mode at a spray voltage of respectively 4350 and 3500 Volts. Full MS scans were between 200 - 1,450 m / z at resolution of 60,000 followed by MS / MS scans of the top 8 most intense ions at a resolution of 30,000. In addition, a target list was added including all the targeted biomarkers. This means that the instrument will prioritize these markers if it cannot detect them it will continue in an untargeted manner.Example 1 - Demonstrating global lipid changes in disease state vs healthy Following LC-MS analysis for candidate ‘diagnostic’ lipids (biomarkers), the following 123 lipid “features” were selected for further analysis;

[0156]

[0157]

[0158] Principal components analysis (PCA) of the candidate biomarkers (in a population of 50 healthy subjects and 50 ALS patients) showed global differences in lipid profiles - see Figure 1a.

[0159] Similarly, PCA of the candidate biomarkers (in a population of 30 healthy subjects and 30 Alzheimer’s Disease patients) showed global differences in lipid profiles - see Figure 1b.

[0160] Similarly, PCA of the candidate biomarkers (in a population of healthy subjects and 8 Parkinson’s Disease patients) showed global differences in lipid profiles - see Figure 1c.

[0161] As can be seen in Figure 1 d, PCA analysis shows that lipid profiles are also distinguished amongst diseases.

[0162] High levels of accuracy are achievable. For example, for distinguishing healthy control (HC) subjects and ALS, the accuracy (on absolutely blinded analysis done here for 2 separate ‘boxes’ of samples) was between 75% and 92% with a combined accuracy of 85.11%. See Figure 1e.

[0163] Example 2 - Machine Learning-Supported Feature Analysis

[0164] This example provides an overall methodology for a machine learning (ML)-supported feature analysis approach, and its application to distinguish disease patients (ALS, AD, PD) from healthy subjects. Following identification of candidate lipids (“features”) based on the initial LC-MS work, the dataset that was subjected to ML analysis included the 123 features outlined in the table above (Example 1).

[0165] The dataset was collected from three different batches, labelled as Box 1, Box 2, and Batch 2. The features contain quantities of lipids present in samples from Healthy Controlsand those diagnosed with either AD, ALS, or PD. These disease groups were used as the targets of our predictive task.

[0166] 51 of these lipids contain non-zero values for at least one sample in Box 1, Box 2, and Batch 2:

[0167]

[0168] Example 2 - Methods

[0169] Cross-validation

[0170] To ensure that the results are reproducible, when running experiments, we used cross- validation to calculate the mean and standard deviation of metrics and feature weights.

[0171] Since some subjects provided multiple samples in our dataset, we split the data into 5 folds, as illustrated in Figure 3, where each person is only represented in either the training or testing split in each fold.

[0172] This ensures that when we are testing the predictive model, it is always acting on samples that are not observed during training.

[0173] When performing pre-processing, all statistics are calculated on the training data only.

[0174] Pre-processing

[0175] Each data point is attributed to one of three batches, which leads to differences in the feature values. This necessitates the use of a method for removing batch variation, which helps to ensure that data across batches is comparable.To do this, we scale and centre all the data from each batch separately, so that the mean and standard deviation of all values for a feature is equal to 0 and 1 respectively.

[0176] This ensures that the data across batches can be compared with each other. For example, in Figure 4, the left-hand plot shows that the data from Batch 2 differs in values from Box 1 and Box 2. After the data is scaled by batch, we see more overlap in the PCA values, illustrated in the right-hand plot.

[0177] We can then apply centring and scaling to the data from all batches together, which is important to ensure that the feature weights of a trained predictive model can be compared.

[0178]

[0179] Feature importance

[0180] The data from all of the batches are organised into two classes, one representing the disease groups AD, ALS, and PD, and the other representing Healthy Controls.

[0181] After the pre-processing is performed, a predictive model (logistic regression with L1 regularisation - see https: / / scikit-learn.org / stable / modules / generated / sklearn.linear model.LogisticRegression.html) is trained on the training portion of the cross-validation fold. From this model, we extract the feature weights, which correspond to their importance. This is done for all of the 5 cross-validation folds, which is additionally repeated 100 times. The distribution of the importance for each feature over the 500 training and testing folds is shown in Figure 5, which distinguishes between features that were positively correlated with a prediction of Healthy Control and positively correlated with AD, ALS, or PD.

[0182] More detail on the methodology to perform feature selection and sample-wise scaling is:

[0183] 1. Remove features that do not contain at least one non-zero value for all batches. 2. For 100 repeats, perform 5-fold validation where we group samples into training and testing splits based on their sample name.

[0184] a. Use the mean and standard deviation of each feature to scale the batches independently. This helps in removing batch effects.*

[0185] b. Use the mean and standard deviation of each feature to scale all of the batches together.

[0186] c. Train a logistic regression model with L2 regularisation.d. Predict on the test folds

[0187] 3. Collect the predictions for all of the 5 folds and calculate the metrics for each of the 100 repeats.

[0188] 4. Calculate the mean and standard deviation of the metrics over the 100 repeats.

[0189] (*where using sample-wise batch scaling, instead of step 2a, can use sample-wise sample scaling on every data point from every batch. This is done by calculating the pairwise difference between features and then dividing that value by the square root of the number of features).

[0190] Over these 500 runs, the features that were, on average, the most important for separating the disease groups from healthy controls were:

[0191] 1. SM 36:2_[M+H]+ (positively correlated with AD, ALS, and PD)

[0192] 2. PC 34:1_[M+H]+ (positively correlated with AD, ALS, and PD)

[0193] 3. PC 36:1_[M+H]+ (positively correlated with AD, ALS, and PD)

[0194] 4. SM 38:2_[M+H]+ (positively correlated with Healthy Control)

[0195] 5. SM 34:2_[M+H]+ (positively correlated with AD, ALS, and PD)

[0196] The full list of ranking (highest to lowest no matter whether positive or negative) and associated feature weight values is below. A positive value is correlated with AD, ALS, and PD; a negative value is correlated with healthy control.

[0197] Feature weights table e.g. for first aspect (ranked 1-51, with SM36:2 being ranked 1 and PC 32:2 being ranked 51):

[0198]

[0199] 1.5869 0.1848 Disease State

[0200] 1.255 0.2455 Disease State

[0201] 1.2122 0.25 Disease State

[0202] -1.1522 0.235 Healthy Control

[0203] 1.1253 0.1708 Disease State

[0204] -1.063 0.2108 Healthy Control

[0205] 0.9882 0.227 Disease State

[0206]

[0207] 0.9519 0.1951 Disease State -0.9283 0.1854 Healthy Control -0.7681 0.2038 Healthy Control -0.7353 0.1965 Healthy Control -0.7293 0.1638 Healthy Control -0.7267 0.1975 Healthy Control 0.6948 0.1978 Disease State -0.6739 0.2352 Healthy Control 0.6212 0.2174 Disease State -0.6156 0.2072 Healthy Control -0.5775 0.1793 Healthy Control 0.5626 0.2173 Disease State 0.5088 0.2094 Disease State 0.5047 0.2542 Disease State -0.4887 0.2127 Healthy Control 0.4786 0.1302 Disease State -0.4428 0.1921 Healthy Control -0.4192 0.179 Healthy Control 0.4078 0.2342 Disease State 0.4054 0.1777 Disease State -0.3895 0.2201 Healthy Control 0.3845 0.2067 Disease State -0.3627 0.2169 Healthy Control 0.3447 0.1685 Disease State 0.3321 0.2208 Disease State -0.3171 0.219 Healthy Control -0.3144 0.1576 Healthy Control -0.301 0.221 Healthy Control 0.2996 0.2406 Disease State 0.2483 0.2849 Disease State 0.2444 0.2486 Disease State 0.2005 0.2433 Disease State 0.1973 0.2001 Disease State 0.1615 0.1894 Disease State -0.1381 0.1398 Healthy Control -0.1307 0.2175 Healthy Control

[0208]

[0209] -0.1214 0.1491 Healthy Control

[0210] -0.103 0.2297 Healthy Control

[0211] -0.0767 0.2491 Healthy Control

[0212] -0.0748 0.2166 Healthy Control

[0213] -0.0606 0.2199 Healthy Control

[0214] 0.0575 0.1989 Disease State

[0215] 0.0451 0.2386 Disease State

[0216] -0.0098 0.215 Healthy Control

[0217]

[0218] Figure 7(a)-(c) outlines the corresponding feature importance / weight rankings when comparing individual diseases vs healthy patients.

[0219] Metrics by number of features

[0220] By only including the top n most important features in our predictive analysis, we can also study how the accuracy, area under the receiver operator characteristic curve (AUC ROC), and Sensitivity vary as the number of features increases. Figure 6 illustrates that the performance of these metrics plateaus as the number of features approaches 15, demonstrating that not all of the 51 features are required to accurately separate the classes.

[0221] Feature p-values forP(Z>\z\):

[0222] By training the same model on all of the training and testing data together, we can calculate the probability that a given feature has a non-zero contribution to the separation of samples labelled as AD, ALS, or PD and Healthy Controls.

[0223] The features that have a p-value of less than 5% are given in Table 1 (logistic regression feature p-values):

[0224]

[0225] Table 1.

[0226] A longer table (beyond the 7 features in the table above) is below, which includes p-values also for comparing individual diseases against healthy subjects. Whether the lipid is increased in disease state vs healthy state (on the one hand) or is increased in healthy state vs disease state (on the other hand) can be seen from the ‘towards’ direction for the corresponding lipid data of Figures 5 and 7. While these p-value data reflect biomarker robustness, it is the “feature importance” approach (e.g. of Figures 5 and 7) that yielded diagnostic lipid biomarkers of principle use in methods of the invention.

[0227]

[0228]

[0229]

[0230]

[0231] Example 3 - Therapy responsiveness

[0232] The inventors had the opportunity to assess (again blindly) serum probes from 3 SOD1 patients taken before treatment with antisense therapy (tofersen) to SOD1 and 6 months after initiating treatment (n=1) and 12 months after initiating treatment (n=2). All 3 pretreatment samples were blindly assessed as ALS. In the 6 months treated patient the sample was still assessed as ‘ALS’ but had almost crossed the border to normal on 2 PCA. In both ALS patients treated for 1 year the samples were assessed as ‘HC’ on 2 PCA (see graph of Figure 8).

[0233]

[0234] This examples is based on data from 112 ALS samples (and 44 control samples) and presents a machine learning framework that identifies highly specific lipid biomarkers capable of distinguishing individuals with amyotrophic lateral sclerosis (ALS) from controls with notably high accuracy. Using multiple models, including Logistic Regression, XGBoost, and Random Forest, trained with nested cross-validation and independent holdout testing, we consistently achieved leading performance (sensitivity: 0.90; specificity: 0.89; AUC-ROC: 0.93). Logistic regression was the most stable and clinically interpretable model, enabling us not only to classify ALS with high sensitivity but also to pinpoint the specific lipid species driving disease signatures. Using SHAP-based feature attribution and model coefficients, we identified a set of highly reproducible ALS-associated lipids, including ether-linked and phosphatidylcholine species such as PCO-34:1, PC 34:1, and PC 34:4. Several sphingomyelins were enriched in controls. This molecular fingerprint reveals a coherent pattern of dysregulated phospholipid remodelling and peroxisomal / lipid-signalling pathways in ALS. By highlighting specific lipid pathways disrupted in ALS, this approach indicates therapeutic targets and pharmacodynamic biomarkers.

[0235] 4.1 Performance evaluation of ALS vs Control classification models

[0236] Model training and evaluation

[0237] To assess predictive model performance, we evaluated multiple classification models using a nested cross validation framework designed to provide an unbiased estimate of generalised performance. Three models, Logistic regression, XGBoost, and Random Forests, were trained and tested using four outer cross-validation folds, with five inner folds used for hyperparameter optimisation. Given the imbalance between classes (44controls and 112 ALS samples), balanced class weighting was applied during model training to ensure balanced treatment of both groups. All analyses in this section are based solely on samples with complete demographic information, ensuring consistency and interpretability across model evaluations. Model performance is outlined in Table 1.

[0238] Table 1: Predictive model performance for classification of ALS vs Controls Model Fl Accuracy Sensitivity Specificity AUCROC XGB 092 (0.89, 096) 088 (0.83, 094) 0.94 (0.91, 0.97) 0.75 (057, 0.93) 0.93 (0.88, 098) RF 0.91 (0.81, 1.00) 0.87 (0.72, 1.00) 0.95 (0.81, 1.00) 0.66 (0.44, 0.88) 092 (0.83, 1.00) LR 0.93 (0.87, 0.98) 0.90 (0.82, 0.97) 0.90 (0.83, 0.97) 0.89 (0.70, 1.00) 0.93 (0.86, 1.00) Each metric is reported as the mean across cross-validation folds with 95% confidence intervals shown in parentheses. Abbreviations: XGB, XGBoost; RF, Random Forests LR, Logistic Regression.

[0239] Model evaluation

[0240] All models achieved F1 scores between 0.91 and 0.93, indicating an excellent balance between precision and sensitivity. Accuracy values ranged from 0.87 to 0.90, reflecting stable correct classification rates despite the relatively small hold-out test sets.

[0241] Sensitivity was consistently high across models (0.90-0.95), indicating reliable detection of ALS cases. In contrast, specificity showed greater variability between models, ranging from 0.66 to 0.89, with tree-based models showing comparatively lower specificity.

[0242] Logistic regression achieved the highest specificity (0.89, 95% Cl: 0.70-1.00), reflecting a stronger ability to correctly classify controls and minimize false-positive ALS predictions.

[0243] The AUC-ROC values ranged from 0.92 to 0.94, confirming consistently strong global discrimination across all thresholds and models. Confidence intervals were tight for most models, indicating stable generalisation across repeated test sets.

[0244] Among the evaluated models, Logistic Regression demonstrated the most consistent performance, the highest generalisability across cross-validation fold, and the most stable performance across evaluation metrics. Based on these properties and its interpretability, Logistic Regression was selected as the primary model for downstream analyses. A confusion matrix outlining the true positive, true negative, false positive, and false negative classifications for the Logistic Regression model is displayed in Figure 10.

[0245] 4.2 Explainability and interpretation of predictionsIn the following section, we present both the Logistic regression coefficients and SHAP values to investigate lipidomic alterations that distinguish ALS from control participants. Coefficients and SHAP values are presented in Figure 11.

[0246] Both methods consistently identified a group of phosphatidylcholine species, particularly PC O 34:1, PC 34:1, LPC 20:4, PC O 36:3, and PC 34:2, as the strongest positive contributors to ALS classification. These lipids exhibited the largest logistic regression coefficients and the highest SHAP impact values, indicating a robust association with ALS in subjects.

[0247] Several sphingomyelin species (e.g., SM 41:1, SM 34:0) showed negative coefficients and negative SHAP contributions, indicating relative enrichment in controls. The concordance between Logistic Regression (LR) coefficients and SHAP values indicates a stable, biologically coherent lipid signature characterised by increased in ALS as outlined in this analysis.

[0248] Without wishing to be bound by theory, it is believed that these patterns align with perturbations in exosomal membrane remodelling, peroxisomal lipid metabolism, and sphingolipid pathways implicated in motor neuron degeneration. Demographic features contributed minimally compared with lipidomic variables. Together, the LR-SHAP feature importance analysis provides both global interpretability and individual-level mechanistic insight, supporting the utility of lipidomics as a discriminative and biologically meaningful source of biomarkers in ALS.

[0249] 4.3 Performance evaluation of ALS vs Control classification models with age and sex imputed

[0250] Imputation of age and sex

[0251] Follow-up experiments were conducted to evaluate whether the model performance reported in Section 3.1 generalises to a larger cohort. While lipidomic biomarker measurements were available for additional participants, age and sex metadata were missing for a subset of these samples, preventing direct use of the same feature set. To address this, we imputed missing demographic values using a maximum-likelihood imputation approach, then re-ran the full training and evaluation workflow on the expanded dataset.To evaluate the impact of demographic imputation on cohort characteristics, we examined age and sex distributions before and after imputation (Figure 12). Imputation was performed using biomarker data alone and did not incorporate diagnostic labels, ensuring that no class-specific information was introduced during this process. Following imputation, the expanded dataset exhibits a more balanced age and sex representation across ALS and control groups, while preserving the overall demographic patterns observed in the original sample. These results indicate that the imputation procedure maintains the underlying demographic structure of the data and supports its use for extending model evaluation to a larger cohort without meaningfully affecting downstream analyses.

[0252] Model evaluation

[0253] Because age and sex contributed only modestly to predictive performance in the original model, imputing these variables is unlikely to materially bias classification results; rather, it enables a consistent evaluation of model robustness on the larger dataset. Predictive model performance is presented in Table 2.

[0254] Compared with performance on the original dataset (Table 1), overall predictive performance on the imputed dataset (Table 2) shows a modest reduction across most metrics, consistent with the increased cohort size and the inclusion of imputed demographic variables. For tree-based models (XGBoostand Random Forest), decreases are observed in F1 score, accuracy, and sensitivity, suggesting some sensitivity to the additional variability introduced through imputation. Despite this, ALICROC values remain relatively stable, indicating preserved overall discriminative ability.

[0255] In contrast, logistic regression demonstrates the greatest robustness to demographic imputation, maintaining strong and well-balanced performance across metrics. While sensitivity is slightly reduced relative to the non-imputed analysis, specificity increases further (0.91 vs. 0.89), and ALICROC improves from 0.93 to 0.95. These results suggest that the logistic regression model generalizes well to the expanded dataset and is less sensitive to uncertainty in imputed demographic variables, reinforcing its suitability as the primary model for downstream analyses. A confusion matrix is provided in Figure 13 outlining the number and type of correctly and incorrectly classified samples.Table 2: Predictive model performance for classification of ALS vs Controls with age and sex imputed. Model Fl Accuracy Sensitivity Specificity AU CROC XGB 0.78 (0.73, 0.83) 082 (0.77, 086) 0.82 (070, 0.94) 0.81 (069, 0.93) 0.91 (086, 0.96) RF 0.71 (0.6, 0.82) 0.78 (0.73, 083) 0.71 (0.47, 0.94) 0.83 (0.70, 0.96) 0.88 (0.83, 0.93) LR 0.84 (0.72, 0.96) 0.88 (0.79. 0.07) 0.82 (0.68, 0.96) 0.81 (0.83, 0.99) 0.95 (0.89, 1.00) Each metric is reported as the mean across cross-validation folds with 95% confidence intervals shown in parentheses. Abbreviations: XGB, XGBoost; RF, Random Forests; LR, Logistic Regression.

[0256] 4.4 Explainability and interpretation of predictions with age and sex imputed

[0257] Compared with the original analysis (Figure 11), the feature importance and interpretability results obtained using the imputed dataset (Figure 14) show a high degree of qualitative consistency, with many of the same lipid classes contributing strongly to ALS-control discrimination. In both analyses, phosphatidylcholine (PC) and ether-linked phosphatidylcholine (PC-O) species remain among the most influential features, exhibiting large positive logistic regression coefficients and consistently positive SHAP values, indicating that elevated levels of these lipids shift model predictions toward ALS.

[0258] Notably, the imputed analysis places greater emphasis on a subset of sphingomyelin (SM) species, including SM 36:2, SM 40:2, and SM 41:1, which show strong and consistent contributions across both coefficients and SHAP values. While the relative ranking of individual features differs modestly between the two analyses, the overall pattern of lipid class involvement is preserved. Demographic variables, including sex, continue to exhibit comparatively small effect sizes in the imputed model, consistent with their limited contribution observed in the original analysis.

[0259] Overall, the close agreement between Figures 11 and 14 indicates that the key drivers of model predictions are stable with respect to demographic imputation. This consistency supports the robustness of the identified lipid signatures and suggests that the primary biological signals underlying ALS classification are not driven by missing demographic information.

[0260] Conclusion

[0261] The key lipid observations (for ALS) are provided in the form of a ranked list of lipid observations in Figure 14C (top to bottom, i.e. where PC 0-34:1 is top ranked, then SM41:1 is second ranked, and so on), where a negative LRC (e.g. pointing to left) indicates whether the lipid is a healthy control observation, and where a positive LRC (e.g. pointing to the right) indicates whether the lipid is an ALS observation. Figure 14C alsoshows the preferred adduct for each key lipid observation for ALS (e.g. the preferred adduct for PC 0-34:1 is [M+H]+).

[0262]

[0263] An analogous workflow to Example 3 above was performed with data for SMA (20 with SMA2; 20 with SMA2) patients. We trained a model using the 38 healthy control samples for whom we have age & sex information and 40 SMA samples. Metrics for the model are: F1: 0.84 (0.74, 0.94)

[0264] Accuracy: 0.85 (0.76, 0.93)

[0265] Precision: 0.88 (0.74, 1.00)

[0266] Sensitivity: 0.82 (0.65, 1.00)

[0267] Specificity: 0.87 (0.72, 1.00)

[0268] AUG ROC: 0.92 (0.81, 1.00)

[0269] Figure 15 outlines a confusion matrix, Logistic Regression weights, and SHAP plot for the SMA model.

[0270] Conclusion

[0271] The key lipid observations (for SMA) are provided in the form of a ranked list of lipid observations in Figure 1D (top to bottom, i.e. where SM 42:2 is top ranked, then SM 40:1 is second ranked, and so on) where a negative LRC (e.g. pointing to left) indicates whether the lipid is a healthy control observation, and where a positive LRC (e.g. pointing to the right) indicates whether the lipid is an SMA observation. Figure 15D also shows the preferred adduct for each key lipid observation for SMA (e.g. the preferred adduct for PC SM 42:2 is [M+H]+).

Claims

CLAIMS1. A method for diagnosing Amyotrophic lateral sclerosis (ALS) in a subject, said method comprisinga) providing a patient lipid profile for said subject,i. wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, andii. wherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;b) providing a reference lipid profile comprising one or more lipid observation(s):i. wherein said reference lipid 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 ALS;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 lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; andiii. wherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:i. when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid observation(s) of the reference lipid profile, said“match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has ALS; and / orii. when the cohort of subjects having ALS is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject has ALS and / or wherein a “non-match” is indicative the subject is healthy;d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said top 5 ranked observations is employed at any one time, said 5 top-ranked observations being selected from:as an ALS observation:PC 0-34:1, PC 34:1 and SM 36:2;as a healthy reference observation:SM 41:1 and PC 32:1;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

2. The method according to claim 1, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: SM 41:1; and ii) as a neurological reference observation: PC 0-34:1 and PC 34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

3. The method according to claim 1, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: SM 41:1 and PC 32:1; and ii) as a neurological reference observation: PC 0-34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

4. The method according to claim 1, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: SM 41:1; and ii) as a neurological reference observation: PC 0-34:1 and SM36:2;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

5. The method according to claim 1, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: PC 32:1; and ii) as a neurological reference observation: PC 0-34:1 and PC 34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

6. The method according to claim 1, wherein the at least three of said 5 topranked observations are:as a neurological reference observation: PC 0-34:1 and PC 34:1 and SM36:2;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

7. The method according to claim 1, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: PC 32:1; and ii) as a neurological reference observation: PC 0-34:1 and SM36:2;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

8. The method according to claim 1, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: SM 41:1 and PC 32:1; and ii) as a neurological reference observation: PC 34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

9. The method according to claim 1, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: SM 41:1; and ii) as a neurological reference observation: PC 34:1 and SM36:2;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

10. The method according to claim 1, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: SM 41:1 and PC 32:1; and ii) as a neurological reference observation: SM36:2;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

11. The method according to claim 1, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: PC 32:1; and ii) as a neurological reference observation: PC 34:1 and SM36:2;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

12. The method according to any preceding claim, wherein at least four of the 5 top-ranked observations are employed, preferably:as an ALS observation: PC 0-34:1, PC 34:1;as a healthy reference observation: SM 41:1 and PC 32:1;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

13. The method according to any preceding claim, wherein at least all five of the 5 top-ranked observations are employed:as an ALS observation: PC 0-34:1, PC 34:1 and SM36:2; andas a healthy reference observation:SM 41:1 and PC 32:1;wherein a “match” with said healthy reference observation and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

14. The method according to any preceding claim, further comprising at least one of the following observations selected from ranked observations 6-10:as an ALS observation:SM 34:2 and LPC 20:4; andas a healthy reference observation:PC 0-36:4, SM 41:2 and SM 34:0;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

15. The method according to claim 14, wherein at least any two of said observations is selected, preferably wherein said at least two observations are selected from:as an ALS observation: SM 34:2; andas a healthy reference observation: PC 0-36:4;wherein a “match” with said healthy reference observation and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

16. The method according to claim 14, wherein at least any two of said observations is selected, preferably wherein said at least any two observations are selected from:i) as a healthy reference observation: SM 41:2; and ii) as an ALS observation: SM34:2;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas an ALS observation: SM34:2 and LPC 20:4;wherein a “match” with said ALS observations is indicative the subject has ALS,whereas a “non-match” with said ALS observations is indicative the subject is healthy;ORi) as a healthy reference observation: SM 34:0; and ii) as an ALS observation: SM34:2;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 0-36:4 and SM 41 :2wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has ALS;ORi) as a healthy reference observation: PC 0-36:4; and ii) as an ALS observation: LPC 20:4;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORas a healthy reference observation: PC 0-36:4 and SM 34:0;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has ALS;ORi) as a healthy reference observation: SM 41:2 ; and ii) as an ALS observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: SM 41:2 and SM 34:0;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has ALS;ORi) as a healthy reference observation: SM 34:0; and ii) as an ALS observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

17. The method according to claim 14, wherein at least any three of said observations ranked 6-10 are selected, preferably wherein said at least three observations are selected from:i) as a healthy reference observation: PC 0-36:4 and SM 41:2; and ii) as an ALS observation: SM34:2;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4; and ii) as an ALS observation: SM34:2 and LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4 and SM 34:0 ; and ii) as an ALS observation: SM34:2;wherein a “match” with said healthy reference observation or a “non-match” with said ALS reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: SM 41:2; ; and ii) as an ALS observation SM34:2 and LPC 20:4:wherein a “match” with said healthy reference observation or a “non-match” with said ALS reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: SM 41:2 and SM 34:0; and ii) as an ALS observation: SM34:2;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: SM 34:0; and ii) as an ALS observation: SM34:2 and LPC 20:4;50wherein a “match” with said healthy reference observation or a “non-match” with said ALS reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4 and SM 41:2 ; and ii) as an ALS observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 0-36:4, SM 41:2 and SM 34:0; wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has ALS;ORi) as a healthy reference observation: PC 0-36:4 and SM 34:0; and ii) as an ALS observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;51i) as a healthy reference observation: SM 41:2 and SM 34:0; and ii) as an ALS observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

18. The method according to claim 14, wherein at least any four of said observations ranked 6-10 are selected, preferably wherein said at least four observations are selected from:i) as a healthy reference observation: SM 41:2, and PC 0-36:4; and ii) as an ALS observation: SM34:2 and LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4, SM 41:2 and SM 34:0; and ii) as an ALS observation: SM34:2;wherein a “match” with said healthy reference observation or a “non-match” with said ALS reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4, and SM 34:0 ; and ii) as an ALS observation: SM34:2 and LPC 20:4;52wherein a “match” with said healthy reference observation or a “non-match” with said ALS reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: SM 41:2 and SM 34:0; and ii) as an ALS observation SM34:2 and LPC 20:4:wherein a “match” with said healthy reference observation or a “non-match” with said ALS reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4, SM 41:2 and SM 34:0; and ii) as an ALS observation: LPC 20:4;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

19. The method according to any preceding claim, wherein at least all five observations selected from ranked observations 6-10 are included, selected from:as an ALS observation: SM 34:2 and LPC 20:4; andas a healthy reference observation: PC 0-36:4, SM 41:2 and SM 34:0; wherein a “match” with said healthy reference observation and a “non-match” with said neurological disorder reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.5320. The method according to any preceding claim, further comprising at least one of the following observations selected from ranked observations 11-15:as an ALS observation: PC 38:3 and SM 42:1; andas a healthy reference observation: SM 40:2, PC 0-38:6 and PC 40:5;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

21. The method according to claim 20, wherein at least two observations from ranked observations 11-15 are included, preferably selected from:i) as a healthy reference observation: SM 40:2; and ii) as an ALS observation: PC 38:3;wherein a “match” with said healthy reference observation or a “nonmatch” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-38:6; and ii) as an ALS observation PC 38:3:wherein a “match” with said healthy reference observation or a “non- match” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORas an ALS observation: PC 38:3 and SM 42:1;54wherein a “match” with said ALS observations is indicative the subject has ALS,whereas a “non-match” with said ALS observations is indicative the subject is healthy;ORi) as a healthy reference observation: PC 40:5; and ii) as an ALS observation PC 38:3:wherein a “match” with said healthy reference observation or a “nonmatch” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: SM 40:2 PC 0-38:6;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has ALS;ORi) as a healthy reference observation: SM 40:2; and ii) as an ALS observation: SM 42:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORas a healthy reference observation: SM 40:2 and PC 40:555wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has ALS;ORi) as a healthy reference observation: PC 0-38:6 ; and ii) as an ALS observation: SM 42:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORas a healthy reference observation: PC 0-38:6 and PC 40:5 wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has ALS;ORi) as a healthy reference observation: PC 40:5; and ii) as an ALS observation: SM 42:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

22. The method according to claim 20, wherein at least three observations from ranked observations 11-15 are included, preferably selected from:i) as a healthy reference observation: SM 40:2 and PC 0-38:6; and ii) as an ALS observation: PC 38:3;56wherein a “match” with said healthy reference observation or a “nonmatch” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: SM 40:2; and ii) as an ALS observation PC 38:3 and SM 42:1:wherein a “match” with said healthy reference observation or a “nonmatch” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: SM 40:2 and PC 40:5; and ii) as an ALS observation: PC 38:3;wherein a “match” with said healthy reference observation or a “non- match” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-38:6; and ii) as an ALS observation PC 38:3 and SM 42:1:wherein a “match” with said healthy reference observation or a “non- match” with said ALS reference observation is indicative the subject is healthy,57whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-38:6, and PC 40:5; and ii) as an ALS observation PC 38:3:wherein a “match” with said healthy reference observation or a “nonmatch” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 40:5; and ii) as an ALS observation: PC 38:3 and SM 42:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: SM 40:2 and PC 0-38:6; and ii) as an ALS observation SM 42:1:wherein a “match” with said healthy reference observation or a “non- match” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;OR58i) as a healthy reference observation: SM 40:2 and PC 0-38:6 ; and ii) as an ALS observation: SM 42:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: SM 40:2 and PC 40:5; ; and ii) as an ALS observation: SM 42:1;wherein a “match” with said healthy reference observation or a “nonmatch” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-38:6 and PC 40:5; and ii) as an ALS observation: SM 42:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

23. The method according to claim 20, wherein at least four observations from ranked observations 11-15 are included, preferably selected from:i) as a healthy reference observation: SM 40:2 and PC 0-38:6; and ii) as an ALS observation: PC 38:3 and SM 42:1;wherein a “match” with said healthy reference observation or a “non- match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;59ORi) as a healthy reference observation: SM 40:2 and PC 0-38:6; and ii) as an ALS observation PC 38:3 and SM 42:1:wherein a “match” with said healthy reference observation or a “nonmatch” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: SM 40:2 and PC 40:5; and ii) as an ALS observation: PC 38:3 and SM 42:1;wherein a “match” with said healthy reference observation or a “nonmatch” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-38:6 and PC 40:5; and ii) as an ALS observation PC 38:3 and SM 42:1:wherein a “match” with said healthy reference observation or a “non- match” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder;OR60i) as a healthy reference observation: SM 40:2, PC 0-38:6, and PC 40:5; and ii) as an ALS SM 42:1;wherein a “match” with said healthy reference observation or a “nonmatch” with said ALS reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has the neurological disorder.

24. The method according to claim 20, wherein at least all five observations from ranked observations 11-15 are included, selected from:as an ALS observation: PC 38:3 and SM 42:1; andas a healthy reference observation:SM 40:2, PC 0-38:6 and PC 40:5;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

25. The method according to any one of claims 20-24, further comprising one or more of the following observations selected from ranked observations 16-35, preferably wherein one or more, two or more, three or more, four or more or all five of observations 16-20 are included, more preferably wherein observations 16-30 are included, and most preferably wherein observations 16-35 are included,wherein said ranked observations 16-35 are, respectively:PC 40:6 (as an ALS observation); SM 38:1 (as a healthy reference observation); SM 33:1 (as an ALS observation); SM 42:2 (as an ALS observation); PC 36:2 (as a healthy reference observation); PC 38:4 (as an ALS observation); SM 32:1 (as an ALS observation); SM 34:1 (as an ALS observation); TG 56:8 (as a healthy reference observation); PC 34:4 (as a healthy reference observation); SM 42:3 (as an ALS observation); TG 50:1 (as an ALS observation); PC 38:4 (as a healthy reference observation); PC 38:6 (as a healthy reference observation); SM 36:1 (as a healthy reference observation); SM 40:1 (as an ALS observation); SM 38:2 (as a healthy reference observation); PC 36:1 (as an ALS observation); PC 0-36:3 (as a healthy reference observation); PC 0-38:5 (as an ALS observation);61wherein a “match” with said healthy reference observation or a “non-match” with said ALS reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said ALS reference observation is indicative the subject has ALS.

26. A method for diagnosing a neurological disorder in a subject, said method comprisinga) providing a patient lipid profile for said subject,i. wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, andii. wherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;b) providing a reference lipid profile comprising one or more lipid observation(s):i. wherein said reference lipid 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 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 lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; andiii. wherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:i. when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient62lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has the neurological disorder; and / orii. when the cohort of subjects having the neurological disorder is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject has the neurological disorder and / or wherein a “non-match” is indicative the subject is healthy;d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said top 5 ranked observations is employed at any one time, said 5 top-ranked observations being selected from:i.as a healthy reference observation:SM 38:2; andii.as a neurological disorder observation:SM 36:2, PC 34:1, PC 36:1 & SM 34:2;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

27. The method according to claim 26, wherein the at least three of said 5 topranked observations are:i) as a neurological disorder observation:63SM 36:2, PC 34:1 & PC 36:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

28. The method according to claim 26, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation:SM 38:2; andii) as a neurological disorder observation:SM 36:2 & PC 34:1;wherein a “match” with said healthy reference observation and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder29. The method according to claim 26, wherein the at least three of said 5 topranked observations are:as a neurological disorder observation:SM 36:2, PC 34:1 & SM 34:2;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

30. The method according to claim 26, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation:SM 38:2; andii) as a neurological disorder observation:SM 36:2 & PC 36:1;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthy,64whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

31. The method according to claim 26, wherein the at least three of said 5 topranked observations are:i) as a neurological disorder observation:SM 36:2, PC 36.1 & SM 34:2;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

32. The method according to claim 26, wherein the at least three of said 5 topranked observations are:(i) as a healthy reference observation:SM 38:2; and(ii) as a neurological disorder observation:SM 36:2 & SM 34:2;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

33. The method according to claim 26, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:SM 38:2; andas a neurological disorder observation:PC 34:1 & PC 36:1;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthy,65whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

34. The method according to claim 26, wherein the at least three of said 5 topranked observations are:as a neurological disorder observation:PC 34:1, PC 36:1 & SM 34:2;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy35. The method according to claim 26, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation:SM 38:2; andii) as a neurological disorder observation:PC 34:1 & SM 34:2;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

36. The method according to claim 26, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation:SM 38:2; andii) as a neurological disorder observation:PC 36:1 & SM 34:2;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthy,66whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

37. The method according to any one of claims 26-36, wherein at least four of the 5 top-ranked observations are employed, preferably:i) as a healthy reference observation:SM 38:2; andii) as a neurological disorder observation:SM 36:2, PC 34:1 & PC 36:1;wherein a “match” with said healthy reference observation and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

38. The method according to any one of claims 26-37, wherein at least all five of the 5 top-ranked observations are employed:i) as a healthy reference observation:SM 38:2; andii) as a neurological disorder observation:SM 36:2, PC 34:1, PC 36:1 & SM 34:2;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said neurological disorder reference observations is indicative the subjectas the neurological disorder.

39. The method according to any one of claims 26-38, further comprising at least one of the following observations selected from ranked observations 6-10:i) as a healthy reference observation:PC 0-36:4 & PC 36:2 & SM 39:1 ; andii) as a neurological disorder observation:PC 0-34:1, & LPC 20:4;67wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

40. The method according to claim 39, wherein at least any two of said observations is selected, preferably wherein said at least two observations are selected from:i) as a healthy reference observation:PC 0-36:4 & SM 39:1; andii) as a neurological disorder observation:any one of PC 0-34: 1,& LPC 20:4;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and a “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation:PC 0-36:4 & PC 36:2wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorder.

41. The method according to claim 39, wherein said at least any two observations are selected from:i) as a healthy reference observation:PC 0-36:2; andii) as a neurological disorder observation:PC 0-34:168wherein a “match” with said healthy reference observation and a “nonmatch” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and a “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4; and ii) as a neurological reference observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 0-36:4 and PC 36:2;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 0-36:4 and SM 39:1;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorder;ORas a neurological reference observation: PC 0-34:1 and LPC 20:4;69wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: PC 36:2; and ii) as a neurological reference observation: PC 0-34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: SM 39:1; and ii) as a neurological reference observation: PC 0-34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 36:2; and ii) as a neurological reference observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;70ORi) as a healthy reference observation: SM 39:1; and ii) as a neurological reference observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 36:2 and SM 39:1;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation:SM 39:1; andii) as a neurological reference observation:PC 0-34:1 & LPC 20:4, or PC 0-34:1;wherein a “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological disorder reference observations is indicative the subject is healthy.

42. The method according to claim 39, wherein at least any three of said observations ranked 6-10 are selected, preferably wherein said at least three observations are selected from:i) as a healthy reference observation:PC 0-36:4; andii) as a neurological disorder observation:71PC 0-34:1 & LPC 20:4wherein a “match” with said healthy reference observation and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and a “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder;ORi) as a healthy reference observation:PC 0-36:4 & PC 36:2; andii) as a neurological disorder observation:LPC 20:4wherein a “match” with said healthy reference observations and a “nonmatch” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation:SM 39:1, PC 0-36:4 & PC 36:2; andwherein a “match” with said healthy reference observations and a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 0-36:4 and PC 36:2; and ii) as a neurological reference observation: PC 0-34:1;72wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4, and SM 39:1; and ii) as a neurological reference observation: PC 0-34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4 and SM 39:1; and ii) as a neurological reference observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 36:2; and ii) as a neurological reference observation: PC 0-34:1 and LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,73whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: SM 39:1; and ii) as a neurological reference observation: PC 0-34:1 and LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 36:2 and SM 39:1; and ii) as a neurological reference observation: PC 0-34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 36:2 and SM 39:1; and ii) as a neurological reference observation: LPC 20:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.7443. The method according to claim 39, wherein at least any four of said observations ranked 6-10 are selected, preferably wherein said at least four observations are selected from:i) as a healthy reference observation:PC 0-36:4 & PC 36:2; andii) as a neurological disorder observation:PC 0-34:1 & LPC 20:4;wherein a “match” with said healthy reference observations and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder;ORi) as a healthy reference observation:SM 39:1, PC 0-36:4 & PC 36:2; andii) as a neurological disorder observation:PC 0-34:1;wherein a “match” with said healthy reference observations and a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder;ORi) as a healthy reference observation:PC 36:2 & SM 39:1; andii) as a neurological disorder observation:PC 0-34:1, & LPC 20:4;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observations is indicative the subject is healthy,75whereas a “non-match” with said healthy reference observation and a “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4 and SM 39:1; and ii) as a neurological reference observation: PC 0-34:1 and LPC 20:4; wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4, PC 36:2 and SM 39:1; and ii) as a neurological reference observation: PC 0-34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

44. The method according to any one of claims 39-43, wherein at least all five observations selected from ranked observations 6-10 are included, selected from:as a healthy reference observation:PC 0-36:4, PC 36:2 & SM 39:1; andas a neurological disorder observation:PC 0-34:1, & LPC 20:4wherein a “match” with said healthy reference observations or a “non- match” with said neurological disorder reference observations is indicative the subject is healthy,76whereas a “non-match” with said healthy reference observations or “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

45. The method according to any one of claims 39-44, further comprising at least one of the following observations selected from ranked observations 11-15:as a healthy reference observation:PC 0-36:3, PC32:1, SM 40:1 & SM 38:1; andas a neurological disorder observation:SM 34:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

46. The method according to claim 45, wherein at least two observations from ranked observations 11-15 are included, selected from:as a healthy reference observation:PC 0-36:3 & PC32:1, PC 0-36:3 & SM 40:1, PC 0-36:3 & SM 38:1 wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations is indicative the subject has the neurological disorder;ORi) as a healthy reference observation:any one of PC 0-36:3, PC32:1, SM 40:1 & SM 38:1; andii) as a neurological disorder observation:SM 34:1;wherein a “match” with said healthy reference observation and a “non- match” with said neurological disorder reference observation is indicative the subject is healthy,77whereas a “non-match” with said healthy reference observation and a “match” with the neurological disorder reference observation is indicative the subject has the neurological disorder.

47. The method according to claim 45, wherein at least three observations from ranked observations 11-15 are included, selected from:as a healthy reference observation:PC 0-36:3, PC 32:1, & SM 40:1 or PC 0-36:3, PC 32:1, & SM 38:1 or PC 0-36:3, SM 40:1 & SM 38:1wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations is indicative the subject has the neurological disorder;ORi) as a healthy reference observation:any two from PC 0-36:3, PC 32:1, SM 40:1 & SM 38:1;for example PC 0-36:3 & PC 32:1, or PC 0-36:3 & SM 40:1, or PC O- 36:3 & SM 38:1, orPC 32:1 & SM 40:1, or PC 32:1 &SM 38:1, orPC 0-36:3 & SM 38:1; andii) as a neurological disorder observation:SM 34:1SM 34:1wherein a “match” with said healthy reference observations and a “non- match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with the neurological disorder reference observation is indicative the subject has the neurological disorder.

48. The method according to claim 45, wherein at least four observations from ranked observations 11-15 are included, selected from:i) as a healthy reference observation:PC 0-36:3, PC 32:1, SM 40:1 & SM 38:178wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations is indicative the subject has the neurological disorder;ORi) as a healthy reference observationPC 0-36:3, PC32:1 & SM 40:1, or PC 0-36:3, PC 32:1 & SM 38:1, or PC 0-36:3, SM 40:1 & SM 38:1ii) as a neurological disorder observation:SM 34:1wherein a “match” with said healthy reference observations and a “nonmatch” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with the neurological disorder reference observation is indicative the subject has the neurological disorder.

49. The method according to claim 45, wherein at least all five observations from ranked observations 11-15 are included, selected from:i) as a healthy reference observation:PC 0-36:3, PC32:1, SM 40:1 & SM 38:1; andii) as a neurological disorder observation:SM 34:1wherein a “match” with said healthy reference observations and a “non- match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

50. The method according to any one of claims 45-49, further comprising one or more of the following observations selected from ranked observations 16-51;preferably wherein one or more, two or more, three or more, four or more or all five of observations 16-20 are included, more preferably79wherein observations 16-35 are included, and most preferably wherein observations 16-51 are included,wherein said ranked observations 16-51 are, respectively:LPC 16:0 (as a neurological reference observation); PC 40:5 (as a healthy reference observation); SM 41:1 (as a healthy reference observation); SM 33:1 (as a neurological reference observation); SM 40:2 (as a neurological reference observation); PC 40:6 (as a neurological reference observation); SM 42:3 (as a healthy reference observation); PC 0-40:8 (as a neurological reference observation); TG 50:1 (as a healthy reference observation); TG 56:8 (as a healthy reference observation); PC 38:6 (as a neurological reference observation); LPC 18:0 (as a neurological reference observation); SM 32:2 (as a healthy reference observation); SM 40:3 (as a neurological reference observation); PC 39:6 (as a healthy reference observation); LPC 20:3 (as a neurological reference observation); LPC 18:0 (as a neurological reference observation); PC 40:4 (as a healthy reference observation); SM 41:2 (as a healthy reference observation); LPC 18:1 (as a healthy reference observation); PC 0-32:0 (as a neurological reference observation); PC 38:3 (as a neurological reference observation); DG 34:1 (as a neurological reference observation); SM 35:1 (as a neurological reference observation); SM 42:1 (as a neurological reference observation); PC 38:4 (as a neurological reference observation); PC 36:2 (as a healthy reference observation); PC 34:4 (as a healthy reference observation); PC 0-38:5 (as a healthy reference observation); PC 0-38:6 (as a healthy reference observation); PC 34:2 (as a healthy reference observation); SM 42:2 (as a healthy reference observation); SM 36:1 (as a healthy reference observation); PC 38:4 (as a neurological reference observation); SM 32:1 (as a neurological reference observation); and PC 32:2 (as a healthy reference observation);wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.8051. The method according to any one of claims 26-50, wherein said neurological disorder comprises ALS, AD & PD.

52. A method for diagnosing Alzheimer’s Disease (AD) in a subject, said method comprisinga) providing a patient lipid profile for said subject,wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, andwherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;b) providing a reference lipid profile comprising one or more lipid observation(s):wherein said reference lipid 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 AD;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 lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; andwherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid81observation(s) of the reference lipid profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has the AD; and / orwhen the cohort of subjects having the AD is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject has the AD and / or wherein a “non-match” is indicative the subject is healthy;d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said 5 topranked observations is employed at any one time, said 5 top-ranked observations being selected from:as a healthy reference observation:SM 38:2, PC36:2, & PC40:4; andas an AD observation:SM34:2 & PC40:5;wherein a “match” with said healthy reference observation and a “non- match” with said AD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD.

53. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:SM 38:2; andas an AD observation:SM 34:2, & PC 40:5;82wherein a “match” with said neurological reference observations is indicative the subject has the AD,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

54. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:SM 38:2 & PC 36:2; andas an AD observation:SM 34:2;wherein a “match” with said healthy reference observation and a “nonmatch” with said AD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD55. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:SM 38:2, & PC 40:4; andas an AD observation:SM 34:2;wherein a “match” with said neurological reference observations is indicative the subject has the AD,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

56. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:SM 38:2, & PC 36:2; andas an AD observation:PC 40:5;83wherein a “match” with said healthy reference observation and a “nonmatch” with said AD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD.

57. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:SM 38:2, PC 36:2, & PC 40:4; andwherein a “match” with said neurological reference observations is indicative the subject has the AD,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy.

58. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:SM 38:2, PC 36:2 & PC40:4; andwherein a “match” with said healthy reference observation and a “non- match” with said AD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD.

59. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 36:2; andas an AD observation:SM 34:2, & PC 40:5;wherein a “match” with said healthy reference observation and a “non- match” with said AD reference observations is indicative the subject is healthy,84whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD.

60. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 40:4; andas an AD observation:SM 34:2, & PC 40:5;wherein a “match” with said neurological reference observations is indicative the subject has the AD,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy61. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 36:2 & PC 40:4; andas an AD observation:SM 34:2;wherein a “match” with said healthy reference observation and a “non- match” with said AD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD.

62. The method according to claim 52, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 36:2, & PC 40:4; andas an AD observation:PC 40:5;wherein a “match” with said healthy reference observation and a “non- match” with said AD reference observations is indicative the subject is healthy,85whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD.

63. The method according to any one of claims 52-62, wherein at least four of the 5 top-ranked observations are employed, preferably:as a healthy reference observation:SM 38:2 & PC36:2; andas an AD observation:SM34:2, & PC40:5;wherein a “match” with said healthy reference observation and a “nonmatch” with said AD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD.

64. The method according to any one of claims 52-63, wherein at least all five of the 5 top-ranked observations are employed:as a healthy reference observation:SM 38:2, PC 36:2 & PC40:4; andas an AD observation:SM 34:2 & PC 40:5;wherein a “match” with said healthy reference observation and a “non- match” with said AD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said AD reference observations is indicative the subject has the AD.

65. The method according to any one of claims 52-64, further comprising at least one of the following observations selected from ranked observations 6-10:as a healthy reference observation:SM 40:2, PC 34:2, PC 0-36:4 & PC 38:4; andas an AD reference observation:PC 38:6;86wherein a “match” with said healthy reference observation or a “non-match” with said AD reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said AD reference observation is indicative the subject has AD.

66. The method according to claim 65, wherein at least any two of said observations ranked 6-10 are selected, preferably wherein said at least two observations are selected from:as a healthy reference observation:any one of SM 40:2, PC 34:2, PC 0-36:4 & PC 38:4; andas an AD reference observation:PC 38:6;wherein a “match” with said healthy reference observation and a “non- match” with said AD reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and a “match” with said AD reference observation is indicative the subject has AD;ORas a healthy reference observation: SM 40:2 and PC 34:2;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorderORas a healthy reference observation:any two of SM 40:2, PC 34:2, PC 0-36:4 & PC 38:4for example SM 40:2 & PC 34:2, or SM 40:2 & PC 0-36:4, or SM 40:2 & PC 38:4, orfor example PC 34:2 & PC 0-36:4 or PC 34:2 & PC 38:4, or for example PC 0-36:4 & PC 38:4wherein a “match” with said healthy reference observations is indicative the subject is healthy,87whereas a “non-match” with said healthy reference observation is indicative the subject has AD.

67. The method according to claim 65, wherein at least any three of said observations ranked 6-10 are selected, preferably wherein said at least three observations are selected from:as a healthy reference observation:any two of SM 40:2, PC 34:2, PC 0-36:4 & PC 38:4for example SM 40:2 & PC 34:2, or SM 40:2 & PC 0-36:4, or SM 40:2 & PC 38:4, orfor example PC 34:2 & PC 0-36:4, or PC 34:2 & PC 38:4, or for example PC 0-36:4 & PC 38:4;as an AD reference observation:PC 38:6;wherein a “match” with said healthy reference observations and a “nonmatch” with said AD reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with said AD reference observation is indicative the subject has AD;ORas a healthy reference observation:any three of SM 40:2, PC 34:2, PC 0-36:4 & PC 38:4for example SM 40:2, PC 34:2 & PC 0-36:4, or SM 40:2, PC 34:2 & PC 38:4, or SM 40:2, PC 0-36:4 & PC 38:4;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations is indicative the subject has AD.

68. The method according to claim 65, wherein at least any four of said observations ranked 6-10 are selected, preferably wherein said at least four observations are selected from:as a healthy reference observation:any three of SM 40:2, PC 34:2, PC 0-36:4 & PC 38:488for example SM 40:2, PC 34:2 & PC 0-36:4, or SM 40:2, PC 34:2 & PC 38:4, or PC 34:2, PC 0-36:4 & PC 38:4;as an AD reference observation:PC 38:6;wherein a “match” with said healthy reference observations and a “nonmatch” with said AD reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with said AD reference observation is indicative the subject has AD;ORas a healthy reference observation:SM 40:2, PC 34:2, PC 0-36:4 & PC 38:4wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations is indicative the subject has AD.

69. The method according to claim 65, wherein at least all five observations selected from ranked observations 6-10 are included, selected from:as a healthy reference observation:SM 40:2, PC 34:2, PC 0-36:4 & PC 38:4; andas an AD reference observation:PC 38:6;wherein a “match” with said healthy reference observations and a “non-match” with said AD reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations and a “match” with said AD reference observation is indicative the subject has AD.

70. The method according to any one of claims 65-69, further comprising at least one of the following observations selected from ranked observations 11-15:as an AD reference observation:PC 36:1, PC 38:4, SM 36:2, PC 0-40:8 & LPC 18:1;89wherein a “match” with said AD reference observation is indicative the subject has AD,whereas a “non-match” with said AD reference observations is indicative the subject is healthy.

71. The method according to claim 70, wherein at least two observations from ranked observations 11-15 are included, selected from:as an AD reference observation:any two of PC 36:1, PC 38:4, SM 36:2, PC 0-40:8 & LPC 18:1, for example PC 36:1 & PC 38:4, or PC 36:1 & SM 36:2, or PC 36:1 & PC 0-40:8, or PC 36:1 & LPC 18:1,for example PC 38:4 & SM 36:2, or PC 38:4 & PC 0-40:8, or PC 38:4 & LPC 18:1,for example SM 36:2 & PC 0-40:8, or SM 36:2 & LPC 18:1, or for example PC 0-40:8 & LPC 18:1;wherein a “match” with said AD reference observations is indicative the subject has AD,whereas a “non-match” with said AD reference observations is indicative the subject is healthy;ORwherein at least three observations from ranked observations 11-15 are included, selected from:as an AD reference observation:any three of PC 36:1, PC 38:4, SM 36:2, PC 0-40:8 & LPC 18:1, for example PC 36:1, PC 38:4 & SM 36:2, or PC 36:1, PC 38:4 & PC O- 40:8, or PC 36:1, PC 38:4 & LPC 18:1, orfor example PC 38:4, SM 36:2 & PC 0-40:8, or PC 38:4, SM 36:2 & LPC 18:1, orfor example SM 36:2, PC 0-40:8 & LPC 18:1;wherein a “match” with said AD reference observations is indicative the subject has AD,whereas a “non-match” with said AD reference observations is indicative the subject is healthy.9072. The method according to claim 70, wherein at least four observations from ranked observations 11-15 are included, selected from:as an AD reference observation:any four of PC 36:1, PC 38:4, SM 36:2, PC 0-40:8 & LPC 18:1, for example PC 36:1, PC 38:4, SM 36:2 & PC 0-40:8, or PC 36:1, PC 38:4, SM 36:2, & LPC 18:1, or PC 36:1, PC 38:4, PC 0-40:8 & LPC 18:1;wherein a “match” with said AD reference observations is indicative the subject has AD,whereas a “non-match” with said healthy reference observations is indicative the subject is healthy;ORwherein at least all five observations from ranked observations 11-15 are included, selected from:as an AD reference observation:PC 36:1, PC 38:4, SM 36:2, PC 0-40:8 & LPC 18:1;wherein a “match” with said AD reference observations is indicative the subject has AD,whereas a “non-match” with said AD reference observations is indicative the subject is healthy.

73. The method according to any one of claims 70-72, further comprising one or more of the following observations selected from ranked observations 16-51, preferably wherein one or more, two or more, three or more, four or more or all five of observations 16-20 are included, more preferably wherein observations 16-35 are included, and most preferably wherein observations 16-51 are included,wherein said ranked observations 16-51 are, respectively:PC 38:3 (as an AD observation); SM 41:2 (as an AD observation); SM 42:2 (as an AD observation); SM 42:1 (as a healthy reference observation); LPC 20:4 (as an AD observation); PC 34:1 (as an AD observation); SM 32:1 (as an AD observation); PC 36:2 (as a healthy reference observation); PC 32:1 (as a healthy reference observation); PC 0-34:1 (as an AD observation); SM 32:2 (as an AD observation); PC 0-36:3 (as a healthy reference observation); SM 36:1 (as an AD observation); SM 35:1 (as a healthy reference observation); PC 40:6 (as91a healthy reference observation); PC 39:6 (as a healthy reference observation); TG 56:8 (as a healthy reference observation); PC 0-32:0 (as an AD observation); LPC 18:0 (as a healthy reference observation); LPC 20:3 (as a healthy reference observation); LPC 16:0 (as an AD observation); SM 41:1 (as a healthy reference observation); LPC 18:0 (as an AD observation); SM 42:3 (as a healthy reference observation); PC 0-38:5 (as a healthy reference observation); SM 34:1 (as an AD observation); SM 33:1 (as an AD observation); SM 40:1 (as a healthy reference observation); SM 40:3 (as an AD observation); TG 50:1 (as a healthy reference observation); SM 39:1 (as an AD observation); SM 38:1 (as a healthy reference observation); PC 32:2 (as an AD observation); PC 0-38:6 (as an AD observation); PC 34:4 (as a healthy reference observation); DG 34:1 (as a healthy reference observation). wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

74. A method for diagnosing Parkinson’s Disease (PD) in a subject, said method comprisinga) providing a patient lipid profile for said subject,wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, andwherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;b) providing a reference lipid profile comprising one or more lipid observation(s):wherein said reference lipid 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 PD;wherein said analysis comprises (a) within each reference cohort, a calculation of a mean expression value and a standard deviation (SD)92value for said one or more lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; andwherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has the PD; and / or when the cohort of subjects having the PD is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject has the PD and / or wherein a “non-match” is indicative the subject is healthy;d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said 5 topranked observations is employed at any one time, said 5 top-ranked observations being selected from:as a healthy reference observation:PC 34:4, PC 36:2, PC 40:4, SM 32:2; and93as a PD observation:SM 40:2;wherein a “match” with said healthy reference observation and a “nonmatch” with said PD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD.

75. The method according to claim 74, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation:PC 34:4, & PC 36:2; andii) as a PD observation:SM 40:2wherein a “match” with said healthy reference observation and a “nonmatch” with said PD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD.

76. The method according to claim 74, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 34:4 & PC 40:4; andas a PD observation:SM 40:2;wherein a “match” with said healthy reference observation and a “non- match” with said PD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD77. The method according to claim 74, wherein the at least three of said 5 topranked observations are:94as a healthy reference observation:PC 34:4 & SM 32:2; andas a PD observation:SM 40:2;wherein a “match” with said healthy reference observations and a “nonmatch” with said PD reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD.

78. The method according to claim 74, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 36:2, & PC 40:4; andas a PD observation:SM 40:2;wherein a “match” with said healthy reference observation and a “nonmatch” with said PD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD.

79. The method according to claim 74, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 36:2, & SM 32:2; andas a PD observation:SM 40:2;wherein a “match” with said healthy reference observations and a “non- match” with said PD reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD.9580. The method according to claim 74, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 40:4, & SM 32:2; andas a PD observation:SM 40:2;wherein a “match” with said healthy reference observation and a “nonmatch” with said PD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD.

81. The method according to claim 74, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 34:4, PC 36:2, & PC 40:4;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorder.

82. The method according to claim 74, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 34:4, PC 36:2 & SM 32:2wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorder.

83. The method according to claim 74, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 34:4, PC 40:4, & SM 32:2; and96wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorder.

84. The method according to claim 74, wherein the at least three of said 5 topranked observations are:as a healthy reference observation:PC 36:2, PC 40:4, & SM 32:2; andwherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorder.

85. The method according to any one of claims 74-84, wherein at least four of the 5 top-ranked observations are employed, preferably:as a healthy reference observation:PC 34:4, PC36:2, & PC40:4; andas a PD observation:SM40:2;wherein a “match” with said healthy reference observation and a “non- match” with said PD reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD.

86. The method according to any one of claims 74-85, wherein at least all five of the 5 top-ranked observations are employed:as a healthy reference observation:PC 34:4, PC36:2, PC40:4, & SM32:2; andas a PD observation:SM40:2;wherein a “match” with said healthy reference observation and a “non- match” with said PD reference observations is indicative the subject is healthy,97whereas a “non-match” with said healthy reference observation and “match” with said PD reference observations is indicative the subject has the PD.

87. The method according to any one of claims 74-86, further comprising at least one of the following observations selected from ranked observations 6-10:as a healthy reference observation:TG56:8 & PC 0-34:1;as a neurological disorder observation:PC38:4, PC38:3, & LPC20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

88. The method according to claim 87, wherein at least any two of said observations is selected, preferably wherein said at least two observations are selected from:i) as a healthy reference observation:TG 56:8; andii) as a neurological disorder observation:PC 38:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

89. The method according to claim 87, wherein at least any two of said observations is selected, preferably wherein said at least any two observations are selected from:i) as a healthy reference observation: TG 56:8; and ii) as a neurological reference observation: PC 38:3;98wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorderORas a healthy reference observation: TG 56:8 and PC 0-34:1;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has the neurological disorder.ORi) as a healthy reference observation: TG 56:8; and ii) as a neurological reference observation: LPC 20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorderORas a neurological reference observation: PC 38:4 and PC 38:3; wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthyOR99i) as a healthy reference observation: PC 0-34:1; and ii) as a neurological reference observation: PC 38:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorderORas a neurological reference observation: PC 38:4 and LPC 20:3; wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthyORi) as a healthy reference observation: PC 0-34:1; and ii) as a neurological reference observation: PC 38:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a neurological reference observation: PC 38:3 and LPC 20:3; wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthyOR100i) as a healthy reference observation: PC 0-34:1 ; and ii) as a neurological reference observation: LPC 20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

90. The method according to claim 87, wherein at least any three of said observations ranked 6-10 are selected, preferably wherein said at least three observations are selected from:i) as a healthy reference observation: TG 56:8; and ii) as a neurological reference observation: PC 38:4 and PC 38:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: TG 56:8 and PC 0-34:1; and ii) as a neurological reference observation: PC 38:4;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: TG 56:8; and ii) as a neurological reference observation: PC 38:4 and LPC 20:3;101wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: TG 56:8 and PC 0-34:1; and ii) as a neurological reference observation: PC 38:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: TG 56:8; and ii) as a neurological reference observation: PC 38:3 and LPC 20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorderORi) as a healthy reference observation: TG 56:8 and PC 0-34:1; and ii) as a neurological reference observation: LPC 20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,102whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-34:1; and ii) as a neurological reference observation: PC 38:4 and PC 38:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a neurological reference observation: PC 38:4 and PC 38:3 and LPC 20:3;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: PC 0-34:1; and ii) as a neurological reference observation: PC 38:4 and LPC 20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;OR103i) as a healthy reference observation: PC 0-34:1; and ii) as a neurological reference observation: PC 38:3 and LPC 20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

91. The method according to claim 87, wherein at least any four of said observations ranked 6-10 are selected, preferably wherein said at least four observations are selected from:i) as a healthy reference observation: TG 56:8 and PC 0-34:1; and ii) as a neurological reference observation: PC 38:4 and PC 38:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: TG 56:8; and ii) as a neurological reference observation: PC 38:4 and PC 38:3 and LPC 20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;OR104i) as a healthy reference observation: TG 56:8 and PC 0-34:1; and ii) as a neurological reference observation: PC 38:4 and LPC 20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-34:1; and ii) as a neurological reference observation: PC 38:3 and LPC 20:3;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-34:1; and ii) as a neurological reference observation: PC 38:4 and PC 38:3 and LPC 20:3; wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

92. The method according to any one of claims 87-91, wherein at least all five observations selected from ranked observations 6-10 are included, selected from:as a healthy reference observation:TG 56:8 and PC 0-34:1; andas a neurological disorder observation:PC 38:4, PC 38:3, LPC 20:3;105wherein a “match” with said healthy reference observations or a “nonmatch” with said neurological disorder reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observations or “match” with said neurological disorder reference observations is indicative the subject has the neurological disorder.

93. The method according to any of claims 87-92, further comprising at least one of the following observations selected from ranked observations 11-15:as a healthy reference observation:LPC 18:1; andas a neurological disorder observation:LPC 18:0, PC 0-40:8, LPC 18:0, & SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

94. The method according to claim 93, wherein at least two observations from ranked observations 11-15 are included, preferably selected from:as a neurological reference observation: LPC 18:0 and PC 0-40:8; wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORas a neurological reference observation: LPC 18:0 and LPC 18:0; wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;106ORas a neurological reference observation: LPC 18:0 and SM 33:1; wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: LPC 18:0;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a neurological reference observation: PC 0-40:8 and LPC 18:0; wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORas a neurological reference observation: PC 0-40:8 and SM 33:1; wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;OR107i) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: PC 0-40:8;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a neurological reference observation: LPC 18:0 and SM 33:1; wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: LPC 18:0;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,108whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

95. The method according to claim 93, wherein at least three observations from ranked observations 11-15 are included, preferably selected from:as a neurological reference observation: LPC 18:0 and PC 0-40:8 and LPC 18:0;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORas a neurological reference observation: LPC 18:0 and PC 0-40:8 and SM 33:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: LPC 18:0 and PC 0-40:8;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;109as a neurological reference observation: LPC 18:0 and LPC 18:0 and SM 33:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: LPC 18:0 and LPC 18:0;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: LPC 18:0 and SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORas a neurological reference observation: PC 0-40:8 and LPC 18:0 and SM 33:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;110ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: PC 0-40:8 and LPC 18:0;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: PC 0-40:8 and SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: LPC 18:0 and SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

96. The method according to claim 93, wherein at least four observations from ranked observations 11-15 are included, preferably selected from:111as a neurological reference observation: LPC 18:0, PC 0-40:8, LPC 18:0 and SM 33:1;wherein a “match” with said neurological reference observations is indicative the subject has the neurological disorder,whereas a “non-match” with said neurological reference observations is indicative the subject is healthy;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: LPC 18:0, PC 0-40:8, LPC 18:0;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: LPC 18:0, PC 0-40:8 and SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: LPC 18:0, SM 33:1, LPC 18:0 ;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,112whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: LPC 18:1; and ii) as a neurological reference observation: PC 0-40:8, LPC 18:0 and SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

97. The method according to claim 93, wherein at least all five observations from ranked observations 11-15 are included, selected fromas a healthy reference observation:LPC 18:1; andas a neurological disorder observation:LPC 18:0, PC 0-40:8, LPC 18:0, & SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder.

98. The method according to any one of claims 93-97, further comprising one or more of the observations selected from ranked observations 16-51, preferably wherein one or more, two or more, three or more, four or more or all five of observations 16-20 are included, more preferably wherein observations 16-35 are included, and most preferably wherein observations 16-51 are included,wherein said ranked observations 16-51 are, respectively:SM 42:2 (as a healthy reference observation); SM 38:2 (as a healthy reference observation); LPC 16:0 (as a PD observation); SM 41:2 (as a healthy reference observation); SM 39:1 (as a healthy reference113observation); PC 40:6 (as a PD observation); PC 34:1 (as a PD observation); SM 40:3 (as a PD observation); PC 36:1 (as a healthy reference observation); TG 50:1 (as a healthy reference observation); SM 42:3 (as a PD observation); PC 32:2 (as a healthy reference observation); SM 40:1 (as a healthy reference observation); SM 41:1 (as a healthy reference observation); PC 0-32:0 (as a healthy reference observation); PC 0-38:6 (as a PD observation); PC 0-36:4 (as a healthy reference observation); SM 42:1 (as a healthy reference observation); PC 0-38:5 (as a healthy reference observation); SM 34:1 (as a PD observation); SM 38:1 (as a PD observation); PC 38:6 (as a PD observation); PC 40:5 (as a PD observation); SM 35:1 (as a healthy reference observation); SM 34:2 (as a healthy reference observation); DG 34:1 (as a PD observation); PC 39:6 (as a healthy reference observation); PC 38:4 (as a PD observation); SM 32:1 (as a healthy reference observation); PC 32:1 (as a healthy reference observation); PC 34:2 (as a PD observation); PC 0-36:3 (as a healthy reference observation); LPC 20:4 (as a healthy reference observation); SM 36:2 (as a healthy reference observation); SM 36:1 (as a PD observation); PC 36:2 (as a healthy reference observation);wherein a “match” with said healthy reference observation or a “non-match” with said neurological disorder reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said neurological disorder reference observation is indicative the subject has the neurological disorder99. A method for diagnosing Spinal Muscular Atrophy (SMA) in a subject, said method comprisinga) providing a patient lipid profile for said subject,i. wherein said patient lipid profile is obtainable by mass spectrometry analysis of the total exosomal lipid extraction content of circulatory extracellular vesicles (EVs) having a diameter of 30-300nm from the subject, andii. wherein said patient lipid profile comprises one or more lipid observation(s), each of which is a different observation;b) providing a reference lipid profile comprising one or more lipid observation(s):114i. wherein said reference lipid 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 SMA;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 lipid observation(s); and (b) between each reference cohort, identification of one or more key lipid observation(s) that demonstrates a mean-fold expression level difference greater than 1.1 or less than -1.1; andiii. wherein said reference lipid profile comprising one or more of said key lipid observation(s) is provided in a form that is suitable for comparative analysis with the patient lipid profile;c) comparing one or more of said key lipid observation(s) present in the reference lipid profile with the corresponding lipid observation(s) present in the patient lipid profile such that:i. when the healthy subject cohort is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) for said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject is healthy and / or wherein a “non-match” is indicative the subject has SMA; and / or ii. when the cohort of subjects having SMA is selected as the reference lipid profile and the mean-fold expression value(s) for one or more of said key lipid observation(s) compared with the corresponding expression value(s) for the same lipid observation(s) of the patient lipid profile, a “match” is defined when the value(s) for one or more key lipid observation(s) of the patient lipid profile falls within 2 SD of the mean expression value(s) of said corresponding one or more key lipid observation(s) of the reference lipid profile, said “match” being indicative the subject has SMA and / or wherein a “non-match” is indicative the subject is healthy;115d) wherein the key lipid observations are provided in the form of a ranked list of lipid observations, the 5 top-ranked key observations being characterised by the presence of at least two phosphatidylcholine observations and at least two sphingomyelin observations, and wherein at least any three of said top 5 ranked observations is employed at any one time, said 5 top-ranked observations being selected from:i.as a healthy reference observation:PC 0-38:6; andii.as a SMA observation:SM 42:2, SM 40:1, SM 36:1, and TG 56:8;wherein a “match” with said healthy reference observation and a “nonmatch” with said SMA reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said SMA reference observations is indicative the subject has SMA.

100. The method according to claim 99, wherein the at least three of said 5 topranked observations are:as an SMA observation: SM 42:2, SM 40:1 and SM 36:1;wherein a “match” with said SMA observations is indicative the subject has SMA, whereas a “non-match” with said SMA observations is indicative the subject is healthy.

101. The method according to claim 99, wherein the at least three of said 5 topranked observations are:as an SMA observation: SM 42:2, SM 40:1 and TG 56:8;wherein a “match” with said SMA observations is indicative the subject has SMA, whereas a “non-match” with said SMA observations is indicative the subject is healthy.

102. The method according to claim 99, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: PC 0-38:6; and ii) as an SMA observation: SM 42:2, SM 40:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,116whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

103. The method according to claim 99, wherein the at least three of said 5 topranked observations are:as an SMA observation: SM 42:2, SM 36:1 and TG 56:8wherein a “match” with said SMA observations is indicative the subject has SMA, whereas a “non-match” with said SMA observations is indicative the subject is healthy.

104. The method according to claim 99, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: PC 0-38:6; and ii) as an SMA observation: SM 42:2, SM 36:1 ;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

105. The method according to claim 99, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: PC 0-38:6; and ii) as an SMA observation: SM 42:2, SM TG 56:8;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

106. The method according to claim 99, wherein the at least three of said 5 topranked observations are:as an SMA observation: SM 40:1, SM 36:1, and TG 56:8 ;wherein a “match” with said SMA observations is indicative the subject has SMA, whereas a “non-match” with said SMA observations is indicative the subject is healthy.117107. The method according to claim 99, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: PC 0-38:6; and ii) as an SMA observation: SM 40:1, SM 36:1 ;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

108. The method according to claim 99, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: PC 0-38:6; and ii) as an SMA observation: SM 40:1, TG 56:8 ;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

109. The method according to claim 99, wherein the at least three of said 5 topranked observations are:i) as a healthy reference observation: PC 0-38:6; and ii) as an SMA observation: SM 36:1, TG 56:8;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

110. The method according to any one of claims 99-109, wherein at least four of the 5 top-ranked observations are employed, preferably:as an SMA observation: SM 42:2, SM 40:1, SM 36:1 and TG 56:8;wherein a “match” with said SMA observations is indicative the subject has SMA, whereas a “non-match” with said SMA observations is indicative the subject is healthy.118111. The method according to any one of claims 99-110, wherein at least all five of the 5 top-ranked observations are employed:as a healthy reference observation: PC 0-38:6; andas a SMA observation: SM 42:2, SM 40:1, SM 36:1, and TG 56:8;wherein a “match” with said healthy reference observation and a “non- match” with said SMA reference observations is indicative the subject is healthywhereas a “non-match” with said healthy reference observation and “match” with said SMA reference observations is indicative the subject has SMA.

112. The method according to any one of claims 99-111, further comprising at least one of the following observations selected from ranked observations 6-10:i) as a healthy reference observation:PC 34:2, PC 32:2, PC 0-36:4 and SM 32:1; andas a SMA observation:PC 0-38:5;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has SMA.

113. The method according to claim 112, wherein at least any two of said observations is selected, preferably wherein said at least two observations are selected from:i) as a healthy reference observation: PC 34:2; andii) as a SMA observation: PC 0-38:5;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has SMA.

114. The method according to claim 112, wherein at least any two of said observations is selected, preferably wherein said at least any two observations are selected from:as a healthy reference observation: PC 34:2 and PC 32:2wherein a “match” with said healthy reference observations is indicative the subject is healthy,119whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 34:2 and PC 0-36:4wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 34:2 and SM 32:1;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORi) as a healthy reference observation: PC 32:2; and ii) as an SMA observation: PC 0-38:5;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 0-36:4; and ii) as an SMA observation: PC 0-38:5;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;120ORi) as a healthy reference observation: SM 32:1; and ii) as an SMA observation: PC 0-38:5;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 32:2 and PC 0-36:4;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 32:2 and SM 32:1;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 0-36:4 and SM 32:1;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;115. The method according to claim 112, wherein at least any three of said observations ranked 6-10 are selected, preferably wherein said at least three observations are selected from:121i) as a healthy reference observation: PC 34:2 and PC 32:2; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 34:2 and PC 0-36:4; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 34:2 and SM 32:1; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 34:2, PC 32:2 and PC 0-36:4; wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;122as a healthy reference observation: PC 34:2, PC 32:2 and SM 32:1; wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 34:2, PC 0-36:4 and SM 32:1; wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORi) as a healthy reference observation: PC 32:2 and PC 0-36:4; ; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 32:2 and SM 32:1; ; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;OR123i) as a healthy reference observation: PC 0-36:4 and SM 32:1; ; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 32:2, PC 0-36:4 and SM 32:1; wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA.

116. The method according to claim 112, wherein at least any four of said observations ranked 6-10 are selected, preferably wherein said at least four observations are selected from:i) as a healthy reference observation: PC 34:2, PC 32:2 and PC 0-36:4; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 34:2, PC 32:2 and SM 32:1; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;124ORi) as a healthy reference observation: PC 34:2, PC 0-36:4 and SM 32:1; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 34:2, PC 32:2, PC 0-36:4 and SM 32:1; wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORi) as a healthy reference observation: PC 32:2, PC 0-36:4 and SM 32:1; and ii) as an SMA observation PC 0-38:5:wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

117. The method according to any one of claims 112-116, wherein at least all five observations selected from ranked observations 6-10 are included, selected from:i) as a healthy reference observation:PC 34:2, PC 32:2, PC 0-36:4 and SM 32:1; andas a SMA observation:PC 0-38:5;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,125whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has SMA.

118. The method according to any one of claims 112-117, further comprising at least one of the following observations selected from ranked observations 11-15: as a SMA observation: SM 33:1; andas a healthy reference observation: PC 38:4, PC 34:4, SM 38:2, and PC 40:5; wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

119. The method according to claim 118, wherein at least two observations from ranked observations 11-15 are included, preferably selected from:i) as a healthy reference observation: PC 38:4; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 34:4; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: SM38:2; and ii) as an SMA observation: SM 33:1;126wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 40:5; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 38:4 and PC 34:4;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 38:4 and SM38:2;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 38:4 and PC 40:5;wherein a “match” with said healthy reference observations is indicative the subject is healthy,127whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 34:4 and SM38:2;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 34:4 and PC 40:5;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: SM38:2 and PC 40:5;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA.

120. The method according to claim 118, wherein at least three observations from ranked observations 11-15 are included, preferably selected from:i) as a healthy reference observation: PC 38:4 and PC 34:4; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;128ORi) as a healthy reference observation: PC 38:4 and PC 34:4; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 38:4 and PC 40:5; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 34:4 and SM38:2; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 34:4 and PC 40:5; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,129whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: SM38:2 and PC 40:5; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 38:4, PC 34:4 and SM38:2; wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 38:4, PC 34:4 and PC 40:5; wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;ORas a healthy reference observation: PC 38:4, SM38:2 and PC 40:5; wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA;130ORas a healthy reference observation: PC 34:4, SM38:2 and PC 40:5;wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA.

121. The method according to claim 118, wherein at least four observations from ranked observations 11-15 are included, preferably selected from:i) as a healthy reference observation: PC 38:4, PC 34:4 and SM38:2; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 38:4, PC 34:4 and PC 40:5; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 38:4, SM38:2 and PC 40:5; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,131whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORi) as a healthy reference observation: PC 34:4, SM38:2 and PC 40:5; and ii) as an SMA observation: SM 33:1;wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder;ORas a healthy reference observation: PC 38:4, PC 34:4, SM38:2 and PC 40:5 wherein a “match” with said healthy reference observations is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation is indicative the subject has SMA.

122. The method according to claim 118, wherein at least all five observations from ranked observations 11-15 are included, selected from:as a SMA observation: SM 33:1; andas a healthy reference observation: PC 38:4, PC 34:4, SM 38:2, and PC 40:5; wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy,whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

123. The method according to any one of claims 118-122, further comprising one or more of the following observations selected from ranked observations 16-34, preferably wherein one or more, two or more, three or more, four or more or all five of132observations 16-20 are included, more preferably wherein observations 16-30 are included, and most preferably wherein observations 16-34 are included,wherein said ranked observations 16-34 are, respectively:SM 41:2 (as a healthy reference observation); PC 36:2 (as a healthy reference observation); PC 0-36:3 (as a healthy reference observation); PC 34:1 (as an SMA observation); LPC 20:4 (as a healthy reference observation); SM 34:1 (as a healthy reference observation); TG 50:1 (as an SMA observation); PC 0-34:1 (as an SMA observation); SM 42:1 (as a healthy reference observation); PC 32:1 as a healthy reference observation); SM 41:1 (as a healthy reference observation); SM 36:2 (as an SMA observation); SM 40:2 (as a healthy reference observation); SM 34:2 (as a healthy reference observation); PC 40:6 (as a healthy reference observation); SM 38:1 (as an SMA observation); SM 34:0 (as a healthy reference observation); PC 38:6 (as an SMA observation); and SM 42:3 (as an SMA observation);wherein a “match” with said healthy reference observation or a “non-match” with said SMA reference observation is indicative the subject is healthy, whereas a “non-match” with said healthy reference observation or “match” with said SMA reference observation is indicative the subject has the neurological disorder.

Images