FIELD OF THE INVENTION
The present invention relates to a novel approach for creating a detailed understanding of toxicological phenomena by elucidating the pathophysiological mechanisms caused by a large number of drugs, compounds, physical, biological and environmental agents.
BACKGROUND AND PRIOR ART OF THE INVENTION
Prior to this, there has been no instance of creating a detailed systems map of a process starlmg with the information obtained from the perturbation of a biological process with several chemical, biological, physical and environmental agents. The traditional methods of constructing a systems map in biological sciences use a serially iterative search of information derived from one of the following:
i. Physiology and associated literature
ii. Clinical measurements and epidemiological observations
iii. Molecular data e.g. toxicogenomie studies
There have been instances of creating disease maps in the recent times, using information
from one or more of the above mentioned alternate strategies. The detailed steps of
constructing systems map (Fig 5) from these alternate approaches are described:
ij
i. Physiology and associated literature - For each biological system, information
on the underlying chemical and molecular reactions have to be obtained from
text books, published hterature or proprietary sources. A serious limitation of
this approach is that it gives all underlying mechanisms equal priority and
does not convey any information about which of these steps would be more
key to the stability of the system. The second limitation of the above approach
is thai one cannot discern non-obvious connections between individual
mechanisms observed in many toxicological studies. Hence one cannot obtain
a holistic picture of the toxicological phenomenon of interest.
ii. Clinical and epidemiological studies - Most of the available data are on
markers of phenomena that are measured as outcomes in clinical studies. They

do not sufficiently convey the mechanistic details of how the system outputs any given marker. Several mechanisms within the system may not uniquely map to a single marker or may not have established markers. A serious limitation of this approach would be that a comprehensive mechanistic map of a system cannot be easily constructed due to lack of appropriate information.
iii. Molecular elucidation from toxicogenomic data - A popular approach in the recent times has been to simultaneously obtain information about thousands of molecules using high-throughput genomic and proteomic studies. The approach would comprise of finding genes or proteins that are perturbed within a system by any given agent, and construction of a map connecting the differentially expressed molecules, based upon experimental observations and published literature-derived information existing in any knowledgebase. There are several limitations of this technique. One major limitation in this approach is that there is no way to prioritize amongst differentially expressed molecules that are primary cornponents of the system and those that are secondary. The second limitation is that in a given experimental setting, only one kind of molecular information (e.g. transcriptional or protein amount or metabolite amount) is available. This requires sophisticated and complex algorithms to integrate data from 'diverse sources and experiments. Experiments are time-consuming and expensive. Information is also confined to a limited number of agents that have been tested with a given technique.
The present invention addresses the following problems:
1. How does one get an overview of all toxic events that a drug or a class of drugs
cause?
2. Are all the toxic events related by some common underlying pathophysiological
mechanisms? I
3. What would be starting point to model such phenomena with a view of predicting
toxic events that can be potentially caused by new chemical entities?

4. If this approach is generalized to represent all perturbing agents such as environmental pollutants/other stressors such as ionizing radiation, it can even provide the systems map of affected biological systems such as ecosystems.
In the present invention, a novel method has been developed to generate a detailed mechanisms map for toxicological phenomena, which is elaborate and yet comprehensive. The core idea is to create a method for generating a detailed mechanisms map for a toxicological phenomenon, using toxic cholestasis as an example.
The invention is able to overcome limitations/problems of previously existing techniques by:
1. Finding more non-obvious yet relevant mechanistic information that can only be obtained from knowledge pertaining to agent-induced perturbations of a system.
2. Using a platform that can easily combine multiple sources (e.g. published literature, high throughput experimental databases that are in public domain or proprietary etc.) of information (molecular, clinical, genomic etc.)
3. Providing consolidated information that can help biological researchers and clinicians get a detailed understanding of the bio-molecules, processes and functions whose derangement causes the toxicological phenomenon of interest.
OBJECTS OF THE INVENTION
The main object of the present invention is to define a high-quality, comprehensive pathophysiological disease map for toxicological phenomena that captures the detailed mechanisms by which various drugs and/or compounds cause the disease and how these mechanisms are interconnected.
Still another object of the present invention is to obtain a method that enables compilation of information about relationships among bio-molecules, and underlying biological mechanisms that are implicated in the disease.

Still another object of the present invention is to develop a new method that utilizes pathway building tools, to provide the starting ground for in-silico predictive modeling of a toxicological phenomenon.
Still another object of the present invention is to provide a method that facilitates comprehensive compilation of information from several sources combined with a semi-automated network building ^nd enrichment analysis that significantly adds to our knowledge base of a given toxicological phenomenon.
Still another object of the present invention is to obtain a method that provides a lead to generate hypothesis for potential biomarkers and targets for drug design in the treatment of any toxicological phenomenon, using a pathophysiological systems map of mechanisms.
Still another object of the present invention is to provide a system for mapping pathophysiological mechanisms by which any drug and/or compound causes a toxicological phenomenon.
Another main object of the present invention is to obtain a novel methodology that streamlines the steps by which' an elaborate and yet comprehensive pathophysiological disease map can be created in a relatively short period of time.
STATEMENT OF THE INVENTION
Accordingly, the present invention relates to a method for mapping pathophysiological mechanisms of agent(s), wherein said method comprising steps of: (i) collecting mechanism terms from plurality of sources; (ii) joining the collected terms into a skeleton map and thereby storing information about the source from which the terms are derived; (iii) searching the agent(s) against specific mechanism terms, and compiling the information in a pathway building tool, to assign each agent(s) to its appropriate mechanism term depending upon the information derived; and (iv) querying molecular interaction database for proteins that interact with the agent(s) and thereafter adding the

proteins into the map; and a :system for mapping pathophysiological mechanisms of agent(s), said system comprises (i) means for collecting mechanism terms from various sources; (ii) pathway analysis tool for searching agent(s) against specific mechanism terms; and (iii) molecular interactions databases for querying proteins that interact with the agent(s).
BRIEF DESCRIPTION OF ACCOMPANYING DRAWINGS
Fig 1: shows the flow chart of the strategy to obtain a pathophysiological map of a toxicological phenomenon, in this case, toxic cholestasis.
Figs 2a, 2b and 2c: show a complete map of toxic Cholestasis, cytoskeleton-related and membrane fluidity related mechanisms respectively. Cytoskeleton and membrane fluidity disruptions etc. are amongst the less described pathophysiological mechanisms that are disrupted by various Cholestatic drugs, which were captured by our method of study in great detail. .
Fig 3: shows Pie chart depicting the percentage of drugs affecting each broadly classified mechanism of toxic cholestasis.
Fig 4: block diagram representaUon of the system.
Fig 5: steps of constructing systems map in the prior art
DETAILED DESCRIPTION OF THE INVENTION
The present invention relates to, a method for mapping pathophysiological mechanisms of agent(s), wherein said method comprising steps of:
i. collecting mechanisrn terms from plurality of sources;
ii. joining the collected terms into a skeleton map and thereby storing information about the source from which the terms are derived ;

iii. searching the agent(s) against specific mechanism terms, and compiling the information in a pathway building tool, to assign each agent(s) to its appropriate mechanism term depending upon the information derived; and
iv. querying molecular interaction database for proteins that interact with the agent(s) and thereafter adding the proteins into the map;
In another embodiment of the present invention, the agent(s) are selected from a group comprising drugs, compounds, physical, biological and environmental agent(s).
A
In yet another embodiment of the present invention, the sources are selected from a group comprising Gene Ontology (GO), Medical Subject Headings (MeSH), textbooks and reviews.
In still another embodiment of the present invention, the proteins relevant to agent(s) are retained either by manual or semi-automated curation.
In still another embodiment of the present invention, the method provides elaborate mechanisms by which individual agent(s) cause a disease.
In still another embodiment of the present invention, the method enables compilation of information about interactions among genes and protein molecules involved in the mechanism for causing the toxicological phenomenon.
In still another embodiment of[ the present invention, the method provides single-click retrieval of references for each mechanism and drug effect.
In still another embodiment of the present invention, the method is useful in in-silico predictive modeling of a disease and provides a lead to generate hypothesis for potential biomarkers and targets for drug design.

In still another embodiment of the present invention, the method is useftil in deriving additional processes and ftinctions related to toxicological phenomena from classifying the proteins added to the pathophysiological mechanisms map.
In still another embodiment of the present invention, the method is used to understand agent induced toxicological phenomena selected from a group comprising loxic Cholestasis, hepatocellular toxicity mechanisms, drug induced renal toxicity, drug induced cardiovascular toxicity^ teratogenicity and drug induced gastrointestinal toxicity, environmental pollutants induced toxicity to ecosystem, preferably, toxic cholestasis.
In still another embodiment of the present invention, the . method provides for comprehensive compilation of information from several sources combined with a semi-automated network building and enrichment analysis and manual curation of output for any pathophysiological disease map.
In still another embodiment of the present invention, the method is useful in running gene ontology classification of the added proteins to provide additional mechanisms to the toxic phenomenon.
The present invention also relates to a system for mapping pathophysiological mechanisms of agent(s), said system comprises
i. means for collecting mechanism terms from various sources; ii. pathway analysis tool for searching agent(s) against specific mechanism terms;
and iii. molecular interactions databases for querying proteins that interact with the agent(s).
In still another embodiment of the present invention, the method is used to understand toxicological phenomena (e.g. toxic cholestasis, hepatocellular toxicity mechanisms, drug-induced renal toxicity, drug-induced cardiovascular toxicity, teratogenicity, drug-

induced gastrointestinal toxicity, environmental pollutants induced toxicity to ecosystem) caused by perturbing agents (e.g. drugs/compounds/metals).
In still another embodiment of the present invention, the method provides elaborate mechanisms by which individual perturbing agents (e.g. drug(s) and/or compound(s)) cause a toxicological phenomena.
In still another embodiment of the present invention, the method enables compilation of

information about interactions

among the underlying mechanisms of the system (e.g.

genes and proteins and processes underlying toxic cholestasis).
In still another embodiment of the present invention, the method provides a map that enables comprehensive storage and retrieval of references for each mechanism and perturbation effect.
In still another embodiment of the present invention, the method has an application in in-silico predictive modeling of a phenomena (e.g. toxic cholestasis).
In still another embodiment of the present invention, the method can be applied to generate hypothesis for potential biomarkers of toxicological phenomena.
In still another embodiment of the present invention, the method provides for comprehensive compilation of information from several sources combined with a semi-automated network building and enrichment analysis and curation of output.
In the present invention, a novel method is developed to generate a detailed mechanisms map for toxicological phenomena, which is elaborate and yet comprehensive. The core idea is to create a method for generating a detailed mechanisms map for a toxicological phenomenon, using toxic cholestasis as an example. Pathway creating tools such as
PathwayArchitcct^'^^ that have! a large information database, allow text mining from the
i) user interface and allow visualization and navigation of big networks allow the user to

custom create interactions, could be suitably adapted for developing our methodology. The method that we have developed allows one to create the pathophysiological systems map of a biological process with sufficient details, exposing lesser known mechanisms which could play key roles in the ftanctioning of the system. The method provides a vantage starting point especially when we are understanding a toxicological phenomena, by directly fmding all basic processes that are effects of the toxicity and the manner in which they are interlinked. This map, could lie at the core of systems modeling of toxicological phenomena.
As a working example, our invention can be used in developing an in-silico predictive model for drug-induced Cholestasis (impairment of bile flow) in the liver. The detailed mechanisms and their associated information presented in the form of a network gives an in-depth understanding of the ibiomolecules, processes and functions and their inter-connectivity, which could be the starting point of model building. We demonstrate the advantages of our approach in the working example of toxic cholestasis by exposing (a). the interconnections amongst the processes that can be deranged and (b) by finding less obvious processes that are likely to be deranged.
This approach could be applied to understand drug-induced toxicological phenomena namely toxic cholestasis, hepatocellular toxicity, drug-induced renal toxicity, drug-induced cardiovascular toxicity, teratogenicity, drug-induced gastrointestinal toxicity. It can also be extended to environmental agent induced disruptions of ecosystems.
The novel methodology developed in this invention starts from studying the toxic effects of hundreds of such agents, elucidates and enriches information about the mechanisms of their actions by analyzing all relevant associated literature and stores the information in an easily retrievable form. This methodology can be drawn to create such pathophysiological maps of any toxicological phenomenon. This new approach of creating a consolidated disease map has drawn from a combination of several high throughput data mining techniques such as text-mining, network building and significance algorithms to give the first comprehensive biological overview of any drug induced toxicological phenomenon.

The method of instant invention comprises the following steps of:
• collecting "terms" suggesting mechanisms underlying the toxicological phenomena from various sources (e.g. mechanisms of bile synthesis, flow; transporters etc from sources such as Gene Ontology (GO), Medical Subject Headings (MeSH), textbooks and a number of reviews on cholestasis).
• creating a preliminary map by putting together the mechanism terms to seed the process of creating a systems map
• storing information about the source from which the terms and their interactions are derived;
• searching individual perturbing agents against specific mechanism terms defining the system
• assigning each perturbing agent to the appropriate mechanism that it deranges and storing the source literature references
• querying a molecular interactions knowledgebase from a pathway analysis tool (e.g. PathwayArchitectTM) for all proteins that interact with the drug(s) and/or compound(s)
• semi-automated curation of the information obtained about the proteins interacting with the drug(s) and/or compounds to retain only those proteins and their interactions that are relevant to the system being studied, in this case, the liver.
• extracting additional mechanisms underlying the system, fi-om the newly added proteins, by using GeneOntology enrichment function.
• compilation of information from several sources combined with a semi-automated network building and enrichment analysis and curation of output.
• incorporating and enriching the prehminary map with the steps detailed above to obtain a comprehensive pathophysiological systems map of a toxicological phenomenon.

The novel method and system (Fig 4) for creating, enriching, compiling and analyzing a
systems map that elucidates the pathophysiological mechanisms of all lexicological
phenomena starting from all chemical, physical, biological and environmental agents
using a reverse engineering approach. Systems maps are created to understand the
If underlying mechanisms that operate and connect together within a biological proccss-
The uniqueness of our approach is in using information mined about the individual agents
to create a mechanistic map of the system that they perturb. The prerequisites for creating
such a pathophysiological systems map are (1) knowledge of several agents that are
known to perturb the system and (2) knowledge of the consequences of each perturbation.
Given these, we demonstrate our new methodology of creating a systems map, by
combining text mining of available information and building a network to demonstrate
the systems operation. We show that our drugs-to-systems approach is superior to the
traditional serial analysis of physiological mechanisms for the same map. Our method has
the potential to extract mechanisms not intuitively obvious as parts of the system, thus
rendering a more comprehensive map.
The new strategy (Fig 1) for^a comprehensive study of a toxicological disorder integrates information from many diverse sources. As a case study, we have used toxic Cholestasis lo demonstrate the efficiency of our methodology. Our strategy involves three basic steps:
1. Collection of large-scale drug-induced pathophysiological data from diverse sources,
2. Deployment of a pathway analysis tool to search, process, store, retrieve and enrich the knowledgebase about the disease.
3. Creating a comprehensive, easily navigable map of the disease based upon the information. The network connections between mechanisms and bio-molecules provide a global overview of the underlying mechanisms and their links as shown in figs 2a, 2b & 2e.

The uniqueness of the map lies in the approach used in elucidating pathophysiological processes by studying the mechanisms of toxicity caused by individual drugs and compounds and compiling all the source information into a single view which can be quickly retrieved. All relevant references and sources of information are stored within the map in a form that is easily retrieved in one click. The use of a pathway database of molecular interactions to add relevant information, in conjunction with the pathway tool's ability to consolidate different types of information, renders the outcome map as a wide coverage and at the same time, compact knowledge base of the disease.
The present invention is further elaborated with the help of following example. However, this example should not be construed to limit the scope of the invention.
WORKING EXAMPLE
The detailed steps of our strategy and how we built the final map for toxic Cholestasis is explained below only as a proof of concept to prove the capabilities of novel methodology used in the present invention.
Obtaining information about detailed mechanisms underlying drug-induced Cholestasis by studying the toxic effects of drugs and compounds on bile flow, elucidates a majority of the pathophysiological mechanisms implicated in the disease.
I WORKFLOW DIAGRAM
■a. Collecting mechanism tenns from GO, MeSH, Zimmermann et al, and reviews.
b. Joining them into a skeleton map and storing information about the source from which
the terms were derived. 1
c. Searching each of the 30;0 known Cholestatic compounds (from Zimmermann et. al.)
against search terms "Cholestasis", "bile", and "liver", in PubMed. The natural language

processing feature of a pathway tool such as, Path way Architect® can be used for this purpose. Tagging of bio-molecules and processes and/or functions facilitated the reading and information extraction. '
d. Assigning each drug to its appropriate mechanism term depending upon the
information derived from the literature search (Fig 3). Table 1 provides a comprehensive
list of drugs associated with each mechanism in the map
e. Searching the molecular interactions database from PA for proteins that interact with
the 300 cholestatic drugs that are placed in the map. Table 2 provides a comprehensive
list of proteins associated with each drug in the map.
Proteins relevant to Cholestasis were retained by curation. 140 proteins were thus added to the map.
APPLICATfONS
The final comprehensive information map for a toxic phenomenon can now be utilized tor several purposes as described below.
I. Mostly, pathophysiological mechanisms of a disease are extracted starting from the features of the disease. This data is scattered in completely different forms in completely different databases and websites. The information in these repositories is limited. Depending on the toxicological phenomenon, we need to develop a creative method to extract maximum information. We have developed one such novel approach to bring in multi-source data into one consoHdated form. For any such phenomenon, the maximum amount of information can be obtained by studying individual agent induced toxic effects. We have utilized this approach to demonstrate how our information compilation covered a much wider range than previously known to exist for toxic Cholestasis.
2. A Systems Biology approach is used to create a global perspective of a disease. This approach essentially requires extensive search and compilation of data from as many

diverse sources as possible and tying them up together into a single, near--complete picture of the phenomenon being studied. Search and compilation are difficult and time-consuming. A variety of computational methods need to be applied in order to carry this out in a short span of time. Here, we have developed a novel method of demonstrating how to use a variety of computational methods and creative approaches to extract maximum information from completely diverse sources and consolidate them into a single, easily searchable ''map" of a toxicological phenomenon. Hence our method is a new approach to search, compile, analyze, store and retrieve multi-source data efficiently and quickly.
3. There is a need to create'composite data formats by amalgamating information from diverse sources and various classes of data. We have shown an effective use of a pathway building tool through this application.
4. Detailed mechanistic information about the involvement of various bio-molecules and processes in a toxic disease provides a lead to generate hypothesis for potential biomarkers and targets for drug design. Using our example of toxic Cholestasis, we can now search for novel biomarkers, which denote the causation, course and response to therapy with respect to the disease. Knowing the mechanisms of Cholestasis in depth will help develop a panel of markers and include novel markers, which would collectively be a better indicator of diagnosis and prognosis of the disease.
'I A long-term application of such a disease map could appear in combination with SNP
(Single Nucleotide Polymorphisms) data analysis. By looking at SNP profiles of individuals and combining the information with a systems map detailing mechanisms of a toxic disease one should be able to predict different individual's responses to drug treatments.
The best way of practicing the invention is in:
o systems biology context of developing the detail mechanisms map of a toxic phenomenon and develop biomarkers for the same

o personalized medicine- Network analysis can be performed based on individual patient data because ^^ariations exist amongst individuals with respect to SNPs, bio-molecules and processes. This will be reflected in their response to various drugs and/or compounds
o in design of modes of intervention to counter the toxic effects of environmental agents on ecosystems.
Advantages of present invention:
(i) Systematic approach ensures completeness of information exploration.
(ii) Semi-automated method expedites the process of analysis, compilation, storage
and retrieval. -S
(iii) More information by analysis and using features of the pathway building
product and its algorithms have enriched the knowledge base of toxicological
I
diseases, (iv) Storage of all information in a comprehensible format within a single view,
which is easily navigable, (v) Utility of the entire comprehensive base of information and analysis on
toxicological diseases, (vi) First comprehensive map of a toxic phenomenon including drugs that cause the
toxicological phenomenon.
It has charted out elaborate mechanisms by which individual perturbing agents cause a toxicological phenomenon, thus giving many more details of mechanisms hitherto poorly defined in the context of the system. We demonstrate this with our working example of toxic cholestasis whereby' our drug-to-disease map revealed the involvement of mechanisms whose combined importance were not intuitively obvious. For example, our analysis on toxic cholestasis revealed that several cholestatic drugs perturbed key mechanisms pertaining to cytoskeietai integrity and membrane fluidity. Thus it is the first instance of reverse engineering a mechanisms map of a toxicological phenomenon.

There are various instances of pathway building tools and text mining being utilized to create a map amongst interacting entities. However, a comprehensive compilation of information from several sources, combined with a semi-automated network building, enrichment analysis and manual curation of final output, has not been reported previously. Obviously, the mix of all these approaches makes our map more detailed and information rich, compared to previously used methods.
Disease maps so far created by other pathway building tools enable only the compilation of information about interactions among genes and protein molecules that underlie normal physiology. None of the maps compiled so far include detailed information related to the various mechanisms implicated in the causation of the disease, f he map created using our proposed strategy and its output is the first exercise towards creating a comprehensive pathophysiological mechanism map, with the associated drugs as well as proteins involved in the disease process that we set as a working example.

We Claim:
t
1. A method for mapping pathophysiological mechanisms of agent(s), wherein said
method comprising steps of:
i. collecting mechanism terms from plurality of sources; ii. joining the collected terms . into a skeleton map and thereby storing
information about the source from which the terms are derived ; iii. searching the agent(s) against specific mechanism terms, and compiling the information in a pathway building tool, to assign each agent(s) to its appropriate mechanism term depending upon the information derived; and iv. querying molecular interaction database for proteins that interact with the agent(s) and thereafter adding the proteins into the map;
2. The method as claimed in claim 1, wherein the agent(s) are selected from a group
comprising drugs, compounds, physical, biological and environmental agent(s)
if and the sources are Selected from a group comprising Gene Ontology (GO),
Medical Subject Headings (MeSH), textbooks and reviews.
3. -The method as claimed in claim 1, wherein the proteins relevant to agent(s) are
retained either by manual or semi-automated curation.
;[
4. The method as clainied in claim 1, wherein the method provides elaborate
mechanisms by which individual agent(s) cause a disease and provides single-
click retrieval of references for each mechanism and drug effect.
f
5. The method as claimed in claim 1, wherein the method enables compilation of
information about interactions among genes and protein molecules involved in the mechanism for causing the toxicological phenomenon.
6. The method as claimed in claim 1, wherein the method is useful in in-silico
predictive modeling of a disease and provides a lead to generate hypothesis for
potential biomarkers and targets for drug design.
26

7. The method as claimed in claim 1, wherein the method is useful in deriving
additional processes and functions related to toxicological phenomena from
classifying the proteins added to the pathophysiological mechanisms map.
i
8. The method as claimed in claim 1, wherein the method is used to understand agent induced toxicological phenomena selected from a group comprising toxic Cholestasis, hepatocellular toxicity mechanisms, drug induced renal toxicity, drug induced cardiovascularftoxieity, teratogenicity and drug induced gastrointestinal toxicity, environmental pollutants induced toxicity to ecosystem, preferably, toxic cholestasis I
9. The method as claimed in claim 1, wherein the method provides for comprehensive compilation of information from several sources combined with a semi-automated network building and enrichment analysis and manual curation of
output for any pathophysiological disease map.
■r
10. The method as claimed in claim 1, wherein the method is useful in running gene
ontology classification'of the added proteins to provide additional mechanisms to
the toxic phenomenon.
i
11. A system for mapping pathophysiological mechanisms of agent(s), said system
comprises
i. means for collecting mechanism terms from various sources;
ii. pathway analysis tool for searching agent(s) against specific mechanism terms;
and
i| iii. molecular interactions databases for querying proteins that interact with the
agent(s).

12. The method and the system for, mapping pathophysiological mechanisms of agent(s) as substantially herein described with reference to accompanying examples and figures.