# US 2026/0100255 A1 · Systems and methods for generating repurposing drug candidates for target diseases

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2026/0100255 A1
- Document ID: US20260100255A1
- Status: Published application
- Kind: A1
- Application number: 19/330,479
- Filed: 2025-09-16
- Published: 2026-04-09
- Earliest claimed: 2024-10-04
- Assignees: Tata Consultancy Services Limited
- Applicants: Tata Consultancy Services Limited
- Inventors: NARAYANAN RAMAMURTHI; SHYAM SUNDAR DAS; PRITISH RANJAN
- Web page: https://patentgrep.com/patent/US20260100255A1

## CPC

- G16C 20/50: Molecular design, e.g. of drugs
- G16C 20/30: Prediction of properties of chemical compounds, compositions or mixtures
- G16C 20/70: Machine learning, data mining or chemometrics
- G16H 70/40: relating to drugs, e.g. their side effects or intended usage
- G16H 20/10: relating to drugs or medications, e.g. for ensuring correct administration to patients
- G16H 50/50: for simulation or modelling of medical disorders
- G16C: COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA

## Abstract

Development of a new drug molecule on average requires an expenditure of US$2-3 billion and a duration of at least 13-15 years. Subsequently, repurposing the existing drugs outside the scope of the original medical indications, has attracted serious consideration. Present disclosure provides a system and method for recommending drug candidates for a target disease by generating drug-refined Medical Subject Headings (MeSH) terms frequency matrix using MeSH terms assigned to Pubmed articles. Further, topic modelling is performed on the drug-refined MeSH terms frequency matrix to generate drug-topic conditional probabilities, and drug-drug distance is computed using Kullback-Leibler divergence. The drugs are then clustered and unknown drugs of a target disease are shortlisted from a cluster with known drugs based on predicted targets and indications of unknown and known drugs. A subset of shortlisted unknown drugs is recommended as a new candidate for target disease based on supporting literature evidence.

## Claims

### Claim 1

A processor implemented method comprising:
obtaining, via one or more hardware processors, an input data pertaining to a plurality of drugs, one or more Medical Subject Headings (MeSH) terms and one or more chemical compounds pertaining to one or more articles, a MeSH tree hierarchy, drug-target mapping information and a disease-target mapping information from a therapeutic targets database (TTD);
processing, via the one or more hardware processors, the input data to obtain a list of matched drugs and a list of refined MeSH terms;
generating, via the one or more hardware processors, a drug-refined MeSH terms frequency matrix based on the list of matched drugs and the list of refined MeSH terms pertaining to the one or more articles;
performing, via the one or more hardware processors, a topic modeling on the drug-refined MeSH terms frequency matrix to obtain one or more drug-topic conditional probabilities of one or more topics comprised of one or more refined MeSH terms amongst the list of refined MeSH terms for each drug amongst the plurality of drugs;
applying, via the one or more hardware processors, a Kullback-Leibler (K-L) divergence on the one or more drug-topic conditional probabilities to obtain a symmetrized pairwise drug-drug distance matrix;
applying, via the one or more hardware processors, one or more clustering techniques on the symmetrized pairwise drug-drug distance matrix to generate one or more drug clusters;
analyzing, via the one or more hardware processors, each cluster amongst the one or more drug clusters to identify at least a subset of clusters containing one or more known drugs specific to a target disease;
segregating, via the one or more hardware processors, a plurality of drugs of each cluster in the identified subset of clusters into a known drug group or unknown drug group;
predicting, for each drug in the known drug group and the unknown drug group, via the one or more hardware processors, one or more targets using an associated chemical information and one or more machine learning algorithms and mapping one or more corresponding indications from the TTD to each of the predicted one or more targets;
performing, via the one or more hardware processors, a comparison of the predicted one or more targets and the corresponding one or more mapped indications of (i) each drug comprised in the unknown drug group with (ii) each drug comprised in the known drug group to obtain a filtered set of unknown drugs having a similarity with one or more known drugs;
identifying, via the one or more hardware processors, one or more literature documents from a literature database for each unknown drugs amongst the filtered set of unknown drugs based on (i) a list of targets relevant to the target disease obtained from the TTD, and (ii) one or more associated PubMed identifiers; and
generating, via the one or more hardware processors, a set of recommended repurposing drug candidates further comprising one or more unknown drugs having the similarity with the one or more known drugs of the target disease based on the one or more literature documents indicating an effectiveness against a specific target relevant to the target disease.

### Claim 2 (depends on 1)

The processor implemented method of claim 1, wherein the one or more drug clusters are generated based on a cluster number representing an associated index.

### Claim 3 (depends on 2)

The processor implemented method of claim 2, wherein the associated index optimizes a cluster stability associated with one or more drug clusters.

### Claim 4

A system comprising:
a memory storing instructions;
one or more communication interfaces; and
one or more hardware processors coupled to the memory via the one or more communication interfaces, wherein the one or more hardware processors are configured by the instructions to:
obtain an input data pertaining to a plurality of drugs, one or more Medical Subject Headings (MeSH) terms and one or more chemical compounds pertaining to one or more articles, a MeSH tree hierarchy, drug-target mapping information and a disease-target mapping information from a therapeutic targets database (TTD);
process the input data to obtain a list of matched drugs and a list of refined MeSH terms;
generate a drug-refined MeSH terms frequency matrix based on the list of matched drugs and the list of refined MeSH terms pertaining to the one or more articles;
perform a topic modeling on the drug-refined MeSH terms frequency matrix to obtain one or more drug-topic conditional probabilities of one or more topics comprised of one or more refined MeSH terms amongst the list of refined MeSH terms for each drug amongst the plurality of drugs;
apply, a Kullback-Leibler (K-L) divergence on the one or more drug-topic conditional probabilities to obtain a symmetrized pairwise drug-drug distance matrix;
apply one or more clustering techniques on the symmetrized pairwise drug-drug distance matrix to generate one or more drug clusters;
analyze each cluster amongst the one or more drug clusters to identify at least a subset of clusters containing one or more known drugs specific to a target disease;
segregate a plurality of drugs of each cluster in the identified subset of clusters into a known drug group or unknown drug group;
predict, for each drug in the known drug group and the unknown drug group, one or more targets using an associated chemical information and one or more machine learning algorithms and mapping one or more corresponding indications from the TTD to each of the predicted one or more targets;
perform a comparison of the predicted one or more targets and the corresponding one or more mapped indications of (i) each drug comprised in the unknown drug group with (ii) each drug comprised in the known drug group to obtain a filtered set of unknown drugs having a similarity with one or more known drugs;
identify one or more literature documents from a literature database for each unknown drugs amongst the filtered set of unknown drugs based on (i) a list of targets relevant to the target disease obtained from the TTD, and (ii) one or more associated PubMed identifiers; and
generate a set of recommended repurposing drug candidates further comprising one or more unknown drugs having the similarity with the one or more known drugs of the target disease based on the one or more literature documents indicating an effectiveness against a specific target relevant to the target disease.

### Claim 5 (depends on 4)

The system of claim 4, wherein the one or more drug clusters are generated based on a cluster number representing an associated index.

### Claim 6 (depends on 5)

The system of claim 5, wherein the associated index optimizes a cluster stability associated with one or more drug clusters.

### Claim 7

One or more non-transitory machine-readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause:
obtaining, an input data pertaining to a plurality of drugs, one or more Medical Subject Headings (MeSH) terms and one or more chemical compounds pertaining to one or more articles, a MeSH tree hierarchy, drug-target mapping information and a disease-target mapping information from a therapeutic targets database (TTD);
processing the input data to obtain a list of matched drugs and a list of refined MeSH terms;
generating a drug-refined MeSH terms frequency matrix based on the list of matched drugs and the list of refined MeSH terms pertaining to the one or more articles;
performing a topic modeling on the drug-refined MeSH terms frequency matrix to obtain one or more drug-topic conditional probabilities of one or more topics comprised of one or more refined MeSH terms amongst the list of refined MeSH terms for each drug amongst the plurality of drugs;
applying a Kullback-Leibler (K-L) divergence on the one or more drug-topic conditional probabilities to obtain a symmetrized pairwise drug-drug distance matrix;
applying one or more clustering techniques on the symmetrized pairwise drug-drug distance matrix to generate one or more drug clusters;
analyzing each cluster amongst the one or more drug clusters to identify at least a subset of clusters containing one or more known drugs specific to a target disease;
segregating a plurality of drugs of each cluster in the identified subset of clusters into a known drug group or unknown drug group;
predicting, for each drug in the known drug group and the unknown drug group, one or more targets using an associated chemical information and one or more machine learning algorithms and mapping one or more corresponding indications from the TTD to each of the predicted one or more targets;
performing a comparison of the predicted one or more targets and the corresponding one or more mapped indications of (i) each drug comprised in the unknown drug group with (ii) each drug comprised in the known drug group to obtain a filtered set of unknown drugs having a similarity with one or more known drugs;
identifying one or more literature documents from a literature database for each unknown drugs amongst the filtered set of unknown drugs based on (i) a list of targets relevant to the target disease obtained from the TTD, and (ii) one or more associated PubMed identifiers; and
generating a set of recommended repurposing drug candidates further comprising one or more unknown drugs having the similarity with the one or more known drugs of the target disease based on the one or more literature documents indicating an effectiveness against a specific target relevant to the target disease.

### Claim 8 (depends on 7)

The one or more non-transitory machine-readable information storage mediums of claim 7, wherein the one or more drug clusters are generated based on a cluster number representing an associated index.

### Claim 9 (depends on 8)

The one or more non-transitory machine-readable information storage mediums of claim 8, wherein the associated index optimizes a cluster stability associated with one or more drug clusters.
