System and method for risk assessment of Parkinsons disease

1 . A method for risk assessment of Parkinson's disease in an individual, the method comprising:

obtaining a sample from a body site of the individual;

extracting Deoxyribonucleic Acid (DNA) from the obtained sample to obtain isolated DNA;

sequencing the isolated DNA using a sequencer to obtain stretches of bacterial DNA sequences;

analyzing, via one or more hardware processors, the stretches of DNA sequences to identify a plurality of bacterial taxa present in the sample, wherein the analysis results in the generation of a bacterial abundance profile having a bacterial abundance value of each of the plurality of bacterial taxa in the sample;

pre-processing, via the one or more hardware processors, the bacterial abundance profile to obtain scaled bacterial abundance values of the bacterial abundance profile;

evaluating, via the one or more hardware processors, a score for each bacterial taxon of the plurality of bacterial taxa for producing a set of neuroactive compounds, wherein the set of neuroactive compounds are compounds which influences the functioning of a gut-brain axis and wherein the score is evaluated independently for each compound of the set of neuroactive compounds and stored in a bacteria-function matrix, wherein the score (SCORBPEO) is calculated using formula:

SCORBPEO

ij

=

P

*

α

*

β

 where P—proportion of strains belonging to the genus ‘j’ that have been predicted with neuroactive compound ‘i’ producing capability, wherein prediction of the neuroactive compound ‘i’ producing capability involves computational identification of enzymes or proteins involved in conversion of tryptophan to the neuroactive compound ‘i’, wherein the identification of enzymes or proteins employ protein sequence or functional domain similarity search algorithms, wherein a filtration step is included based on presence of genes or functional domains of a particular pathway in proximity to each other in a genome of a particular organism to increase prediction efficiency,

 α—confidence value of the corresponding bacterial group, where the confidence value is evaluated based on the relative number of strains belonging to a particular genus, and

 β—‘weightage’ which represents an enrichment value of a particular pathway in a particular body site;

calculating, via the one or more hardware processors, a metabolic potential (MP) corresponding to each compound of the set of neuroactive compounds using the bacteria function matrix and the scaled bacterial abundance values, wherein the metabolic potential (MP) is indicative of the capability of the bacterial community for producing the neuroactive compound, wherein the metabolic potential (MP) is calculated using formula:

MP

NAC

=

∑

i

=

1

n

RA

i

×

SCORBPEO

[

NAC

]

[

i

]

 where, MP NAC —Metabolic potential of the bacterial community (of interest) for production of a particular neuroactive compound,

 n—number of the particular neuroactive compound producing bacterial genera present in the bacterial community of interest,

 RA—relative scaled abundance of a particular bacterial genus ‘i’ predicted to have the metabolic pathway for the neuroactive compound production, and

 SCORBPEO [NAC] [i] —The ‘SCORBPEO (Score for Bacterial Production of Neuro-active Compound)’ score of genus ‘i’ for production of the particular neuroactive compound ‘NAC’;

generating, via the one or more hardware processors, a binary classification model utilizing the metabolic potential (MP) of each compound of the set of neuroactive compounds using machine learning techniques, wherein MP values corresponding to the set of neuroactive compounds are considered as features for training the binary classification model, wherein a sample set is randomly divided into training set samples and testing set samples, in a ratio such that a proportion of diseased and healthy samples remained equivalent in both training and test set, wherein the training of the binary classification model is performed by utilizing a random forest package with 10-fold cross-validation, wherein from each cross-validation fold, differentiating features are picked and a giniscore is utilized to rank the differentiating features based upon their cumulative importance, and the ranked differentiating features are progressively added into the binary classification model, based on their cumulative importance, wherein performance of the binary classification model is evaluated after addition of every new feature;

predicting, via the one or more hardware processors, the risk of the individual to develop or suffering from Parkinson's disease in a risk or no risk, using the binary classification model based on a predefined set of conditions, wherein the individual has the risk;

designing therapeutic approaches, through targeting the bacterial groups that are capable of producing a set of neurotoxic compounds or facilitating growth of healthy microbes, wherein the set of neurotoxic compounds are compounds which negatively affect the functioning of the gut-brain axis, wherein designing the therapeutic approaches includes identification of a consortium of microbes to be used in the form of prebiotic or probiotic or symbiotic to (i) reduce growth of bacteria in the gut which are capable of producing the set of neurotoxic compounds and (ii) enhance the growth of beneficial bacteria in the gut which is capable of maintaining a healthy gut or produce the set neuroactive compounds which are beneficial for functioning and regulation of the gut-brain axis,

wherein the identification of the consortium of microbes is performed based on the MP values of a bacterial genera identified in the sample, differentially abundant bacterial taxa in cohorts of Parkinson's disease patients and healthy individuals and the SCORBPEO score values of the differentially abundant bacterial taxa corresponding to the production of the set of neuroactive compounds; and

administering, to the individual with the risk, the designed consortium of microbes for improved treatment of Parkinson's disease, including minimization of side effects of therapeutic drugs through maintaining a healthy cohort of the bacteria in the gut,

wherein the method is implemented for diagnosis or treatment of Parkinson's disease.

2 . The method according to claim 1 , wherein the predefined set of conditions comprises comparing the metabolic potential for production of one of the set of neuroactive compounds with a threshold value, wherein the result of comparison is:

no risk of Parkinson's disease if the metabolic potential is less than or equal to the threshold value, or

a significant risk of Parkinson's disease if the metabolic potential is more than the threshold value.

3 . The method according to claim 1 , wherein isolating and sequencing stretches of DNA further comprises at least one of:

amplifying and sequencing bacterial 16S rRNA, 23S rRNA, rpoB, or cpn60 marker genes from the bacterial DNA,

amplifying and sequencing one or more of a full-length or one or more specific regions of the bacterial 16S rRNA, 23S rRNA, rpoB, cpn60 marker genes from the microbial DNA,

amplifying and sequencing one or more phylogenetic marker genes from the bacterial DNA, or

whole genome shotgun sequencing (WGS) data corresponding to bacterial DNA, isolated from the body site of the individual.

4 . The method according to claim 1 , wherein the step of sequencing comprises one or more of: an amplicon sequencing, a whole genome shotgun sequencing (WGS), a fragment library based sequencing technique, a mate-pair library or a paired-end library based sequencing technique, a polymerase chain reaction (PCR), an RNA sequencing technique, or a microarray-based technique.

5 . The method according to claim 1 , wherein the step of pre-processing the microbial abundance data comprises normalizing to represent the abundance in the form of scaled values, wherein the normalization on microbial counts is performed through one or more of a rarefaction, a quantile scaling, a percentile scaling, a cumulative sum scaling, or an Aitchison's log-ratio transformation.

6 . The method according to claim 1 wherein the set of neuroactive compounds comprises one or more of Kynurenine, Quinolinate, Indole, Indole acetic acid (IAA), Indole propionic acid (IPA), Tryptamine, Gamma-amino butyric acid (GABA), and Hydrogen sulphide (H 2 S).

7 . The method according to claim 1 , wherein generating the binary classification model using machine learning techniques is performed using one or more of random forest, decision trees techniques, linear regression, logistic regression, naive Bayes, linear discriminant analyses, k-nearest neighbour algorithm, Support Vector Machines, and Neural Networks techniques.

8 . The method according to claim 1 , wherein the sample is one of saliva, stool, blood, body fluid, tissue, or swab.

9 . The method according to claim 1 , wherein the body site is one of a gut, oral, skin, or urinogenital tract of the individual.

10 . The method according to claim 1 , wherein the healthy microbes include microbes producing neuro-protective compounds which have beneficial effects on the gut-brain axis.

11 . The method according to claim 1 , wherein values of the computed score is rescaled to 0-10, wherein for the particular pathway, a bacterial taxa having a higher SCORBPEO indicate a greater probability of production of a particular neuroactive compound as compared to a taxa with a lower SCORBPEO, wherein scaling of the bacterial abundance values is performed through one or more of: a percentage, a cumulative sum scaling, a minmax scaling, a maxAbs scaling, a robust scaling, a percentile, a quantile, or an Atkinson's log transformation, wherein percentage scaled bacterial abundance data is pre-processed to remove bacterial genera having missing or null values in the samples analyzed to cleanse the bacterial abundance data and remove inconsistencies in the bacterial abundance data before generating the binary classification model and the pre-processed data is transformed by scaling from 0 to 1 such that the abundances of the bacterial genera in different samples are placed at a common scale.

12 . The method according to claim 1 , wherein the consortium of microbes is composed of a combination of GABA and Indole to classify healthy samples and diseased samples, wherein the bacterial strains belonging to genera Roseburia and Butyricicoccus are administered as probiotic formulation in order to alleviate symptoms of Parkinson's disease through maintaining a healthier gut microbiome, thereby improving an efficiency of current therapeutic approaches or minimizing the side effects of conventional therapies of Parkinson's disease.